Ciencia y tecnología

Massive, superficial AI-generated content isn't just a problem, it's also a symptom. Technology amplifies a consumption model that rewards fluidity and drains our attention span.

We listen to interviews, podcasts and audios of our family at 2x. We watch videos cut into highlights, and we base decisions and criteria on articles and reports that we have only read summarized with AI. We consume information in ultra-fast mode, but at a cognitive level we give it the same validity as when we consumed it more slowly, and we even apply it in decision-making. What is affected by this process is not the basic memory of contents, which seems to be maintained according to controlled studies, but the ability to connect that knowledge with what we already had and to elaborate our own ideas with it. More than superficiality, it is worrying that this new way of thinking is sufficient in so many contexts.

What's new and what's not?

We may think that generative AI has only intensified an old dynamic in which content production is infinite, but our attention spans are the same. We cannot fool ourselves either, because since the Internet has existed, infinity is not new. If we were to say that the problem is that there is too much content, we would be complaining about a situation in which we have been living for more than twenty years. Nor is the crisis of authority of official information or the difficulty in distinguishing reliable sources from those that are not.

However, the AI slop, which is the flood of AI-generated digital content on the Internet, brings its own logic and new considerations, such as breaking the link between effort and content, or that all that is generated is a statistical average of what already existed. This uniform and uncontrolled flow has consequences: behind the mass-generated content there may be an orchestrated intention of manipulation, an algorithmic bias, voluntary or not, that harms certain groups or slows down social advances, and also a random and unpredictable distortion of reality.

 But how much of what I read is AI?

By 2025, it has been estimated that a large portion of online content  incorporates synthetic text: an Ahrefs analysis of nearly one million web  pages published in the first half of the year found that 74.2% of new pages contained signals of AI-generated content. Graphite research from the same year cites that, during the first year of ChatGPT alone, 39% of all online  content was already generated with AI. Since November 2024, that figure has remained stable at around 52%, meaning that since then AI content outnumbers human content.

However, there are two questions we should ask ourselves when we come across estimates of this type:

1. Is there a reliable mechanism to distinguish a written text from a generated text? If the answer is no, no matter how striking and coherent the conclusions are, we cannot give them value, because they could be true or not. It is a valuable quantitative data, but one that does not yet exist.

With the information we currently have, we can say that "AI-generated text" detectors fail as often as a random model would, so we cannot attribute reliability to them. In a recent study cited by The Guardian, detectors were correct about whether the text was generated with AI or not in less than 40% of cases. On the other hand, in the first paragraph of Don Quixote, certain detectors have also returned an 86% probability that the text was created by AI.

2. What does it mean that a text is generated with AI? On the other hand, the process is not always completely automatic (what we call copying and pasting) but there are many grays in the scale: AI inspires, organizes, assists, rewrites or expands ideas, and denying, delegitimizing or penalizing this writing would be ignoring an installed reality.

The two nuances above do not cancel out the fact that the AI slop exists, but this does not have to be an inevitable fate. There are ways to mitigate its effects on our abilities.

What are the antidotes?

We may not contribute to the production of synthetic content, but we cannot slow down what is happening, so the challenge is  to review the criteria and habits of mind with which we approach both reading and writing content.

1. Prioritize what clicks: one of the few reliable signals we have left is that clicking sensation at the moment when something connects with a previous knowledge, an intuition that we had diffused or an experience of our own, and reorganizes it or makes it clear. We also often say that it "resonates". If something clicks, it's worth following, confirming, researching, and briefly elaborating on a personal level.

2. Look for friction with data: anchoring content in open data and verifiable sources introduces healthy friction against the AI slop. It reduces, above all, arbitrariness and the feeling of interchangeable content, because the data force us to interpret and put it in context. It is a way of putting stones in the excessively fluid river that is the generation of language, and it works when we read and when we write.

3. Who is responsible? The text exists easily now, the question is why it exists or what it wants to achieve, and who is ultimately responsible for that goal. It seeks the signature of people or organizations, not so much for authorship but for responsibility. He is wary of collective signatures, also in translations and adaptations.

4. Change the focus of merit: evaluate your inertia when reading, because perhaps one day you learned to give merit to texts that sounded convincing, used certain structures or went up to a specific register. It shifts value to non-generatable elements such as finding a good story, knowing how to formulate a vague idea or daring to give a point of view in a controversial context.

On the other side of the coin, it is also a fact that content created with AI enters with an advantage in the flow, but with a disadvantage in credibility. This means that the real risk now is that AI can create high-value content, but people have lost the ability to concentrate on valuing it. To this we must add the installed prejudice that, if it is with AI, it is not valid content. Protecting our cognitive abilities and learning to differentiate between compressible and non-compressible content is therefore not a nostalgic gesture, but a skill that in the long run can improve the quality of public debate and the substrate of common knowledge.

Content created by Carmen Torrijos, expert in AI applied to language and communication. The content and views expressed in this publication are the sole responsibility of the author.

Open data is a central piece of digital innovation around artificial intelligence as it allows, among other things, to train models or evaluate machine learning algorithms. But between "downloading a CSV from a portal" and accessing a dataset ready to apply machine learning techniques , there is still an abyss.

Much of that chasm has to do with metadata, i.e. how datasets are described (at what level of detail and by what standards). If metadata is limited to title, description, and license, the work of understanding and preparing data becomes more complex and tedious for the person designing the machine learning model. If, on the other hand, standards that facilitate interoperability are used, such as DCAT, the data becomes more FAIR (Findable, Accessible, Interoperable, Reusable) and, therefore, easier to reuse. However, additional metadata is needed  to make the data easier to integrate into machine learning flows.

This article provides an overview of the various initiatives and standards needed to provide open data with metadata that is useful for the application of machine learning techniques.

DCAT as the backbone of open data portals

The DCAT (Data Catalog Vocabulary) vocabulary was designed by the W3C to facilitate interoperability between data catalogs published on the Web. It describes catalogs, datasets, and distributions, being the foundation on which many open data portals are built.

In Europe, DCAT is embodied in the DCAT-AP application profile, recommended by the European Commission and widely adopted to describe datasets in the public sector, for example, in Spain with DCAT-AP-ES. DCAT-AP answers questions such as:

• What datasets exist on a particular topic?
• Who publishes them, under what license and in what formats?
• Where are the download URLs or access APIs?

Using a standard like DCAT is imperative for discovering datasets, but you need to go a step further in order to understand how they are used in machine learning models or what quality they are from the perspective of these models.

MLDCAT-AP: Machine Learning in an Open Data Portal Catalog

MLDCAT-AP (Machine Learning DCAT-AP) is a DCAT application profile developed by SEMIC and the Interoperable Europe community, in collaboration with OpenML, that extends DCAT-AP to the machine learning domain.

MLDCAT-AP incorporates classes and properties to describe:

• Machine learning models and their characteristics.
• Datasets used in training and assessment.
• Quality metrics obtained on datasets.
• Publications and documentation associated with machine learning models.
• Concepts related to risk, transparency and compliance with the European regulatory context of the AI Act.

With this, a catalogue based on MLDCAT-AP no longer only responds to "what data is there", but also to:

• Which models have been trained on this dataset?
• How has that model performed by certain metrics?
• Where is this work described (scientific articles, documentation, etc.)?

MLDCAT-AP represents a breakthrough in traceability and governance, but the definition of metadata is maintained at a level that does not yet consider the internal structure of the datasets or what exactly their fields mean. To do this, it is necessary to go down to the level of the structure of the dataset distribution itself.

Metadata at the internal structure level of the dataset

When you want to describe what's inside the distributions of datasets (fields, types, constraints), an interesting initiative is Data Package, part of the Frictionless Data ecosystem.

A Data Package is defined by a JSON file that describes a set of data. This file includes not only general metadata (such as name, title, description or license) and resources (i.e. data files with their path or a URL to access their corresponding service), but also defines a schema with:

• Field names.
• Data types (integer, number, string, date, etc.).
• Constraints, such as ranges of valid values, primary and foreign keys, and so on.

From a machine learning perspective, this translates into the possibility of performing automatic structural validation before using the data. In addition, it also allows for accurate documentation of the internal structure of each dataset and easier sharing and versioning of datasets.

In short, while MLDCAT-AP indicates which datasets exist and how they fit into the realm of machine learning models, Data Package specifies exactly "what's there" within datasets.

Croissant: Metadata that prepares open data for machine learning

Even with the support of MLDCAT-AP and Data Package, it would be necessary to connect the underlying concepts in both initiatives. On the one hand, the field of machine learning (MLDCAT-AP) and on the other hand, that of the internal structures of the data itself (Data Package). In other words, the metadata of MLDCAT-AP and Data Package may be used, but in order to overcome some limitations that both suffer, it is necessary to complement it. This is where Croissant comes into play, a metadata format for preparing datasets for machine learning application. Croissant is developed within the framework of MLCommons, with the participation of industry and academia.

Specifically, Croissant is implemented in JSON-LD and built on top of schema.org/Dataset, a vocabulary for describing datasets on the Web. Croissant combines the following metadata:

• General metadata of the dataset.
• Description of resources (files, tables, etc.).
• Data structure.
• Semantic layer on machine learning (separation of training/validation/test data, target fields, etc.)

It should be noted that Croissant is designed so that different repositories (such as Kaggle, HuggingFace, etc.) can publish datasets in a format that machine learning libraries (TensorFlow, PyTorch, etc.) can load homogeneously. There is also a CKAN extension to use Croissant in open data portals.

Other complementary initiatives

It is worth briefly mentioning other interesting initiatives related to the possibility of having metadata to prepare datasets for the application of machine learning ("ML-ready datasets"):

• schema.org/Dataset: Used in web pages and repositories to describe datasets. It is the foundation on which Croissant rests and is integrated, for example, into Google's structured data guidelines to improve the localization of datasets in search engines.
• CSV on the Web (CSVW): W3C set of recommendations to accompany CSV files with JSON metadata (including data dictionaries), very aligned with the needs of tabular data documentation that is then used in machine learning.
• Datasheets for Datasets and Dataset Cards: Initiatives that enable the development  of narrative and structured documentation to describe the context, provenance, and limitations of datasets. These initiatives are widely adopted on platforms such as Hugging Face.

Conclusions

There are several initiatives that help to make a suitable metadata definition for the use of machine learning with open data:

• DCAT-AP and MLDCAT-AP articulate catalog-level, machine learning models, and metrics.
• Data Package describes and validates the structure and constraints of data at the resource and field level.
• Croissant connects this metadata to the machine learning flow, describing how the datasets are concrete examples for each model.
• Initiatives such as CSVW or Dataset Cards complement the previous ones and are widely used on platforms such as HuggingFace.

These initiatives can be used in combination. In fact, if adopted together, open data is  transformed from simply "downloadable files" to machine learning-ready raw material, reducing friction, improving quality, and increasing trust in AI systems built on top of it.

Jose Norberto Mazón, Professor of Computer Languages and Systems at the University of Alicante. The contents and views expressed in this publication are the sole responsibility of the author.

Three years after the acceleration of the massive deployment of Artificial Intelligence began with the launch of ChatGPT, a new term emerges strongly: Agentic AI. In the last three years, we have gone from talking about language models (such as LLMs) and chatbots (or conversational assistants) to designing the first systems capable not only of answering our questions, but also of acting autonomously to achieve objectives, combining data, tools and collaborations with other AI agents or with humans. That is, the global conversation about AI is moving from the ability to "converse" to the ability to "act" of these systems.

In the private sector, recent reports from large consulting firms describe AI agents that resolve customer incidents from start to finish, orchestrate supply chains, optimize inventories in the retail sector  or automate business reporting. In the public sector, this conversation is also beginning to take shape and more and more administrations are exploring how these systems can help simplify procedures or improve citizen service. However, the deployment seems to be somewhat slower because logically the administration must not only take into account technical excellence but also strict compliance with the regulatory framework, which in Europe is set by the AI Regulation, so that autonomous agents are, above all, allies of citizens.

What is Agentic AI?

Although it is a recent concept that is still evolving, several administrations and bodies are beginning to converge on a definition. For example, the UK government describes agent AI as systems made up of AI agents that "can autonomously behave and interact to achieve their goals." In this context, an AI agent would be a specialized piece of software that can make decisions and operate cooperatively or independently to achieve the system's goals.

We might think, for example, of an AI agent in a local government who receives a request from a person to open a small business. The agent, designed in accordance with the corresponding administrative procedure, would check the applicable regulations, consult urban planning and economic activity data, verify requirements, fill in draft documents, propose appointments or complementary procedures and prepare a summary so that the civil servants could review and validate the application. That is, it would not replace the human decision, but would automate a large part of the work between the request made by the citizen and the resolution issued by the administration.

Compared to a conversational chatbot – which answers a question and, in general, ends the interaction there – an AI agent can chain multiple actions, review results, correct errors, collaborate with other AI agents and continue to iterate until it reaches the goal that has been defined for it. This does not mean that autonomous agents decide on their own without supervision, but that they can take over a good part of the task always following well-defined rules and safeguards.

Key characteristics of a freelance agent include:

• Perception and reasoning: is the ability of an agent to understand a complex request, interpret the context, and break down the problem into logical steps that lead to solving it.
• Planning and action: it is the ability to order these steps, decide the sequence in which they are going to be executed, and adapt the plan when the data changes or new constraints appear.
• Use of tools: An agent can, for example, connect to various APIs, query databases, open data catalogs, open and read documents, or send emails as required by the tasks they are trying to solve.
• Memory and context: is the ability of the agent to maintain the memory of interactions in long processes, remembering past actions and responses and the current state of the request it is resolving.
• Supervised autonomy: an agent can make decisions within previously established limits to advance towards the goal without the need for human intervention at each step, but always allowing the review and traceability of decisions.

We could summarize the change it entails with the following analogy: if LLMs are the engine of reasoning, AI agents are systems that , in addition to the ability to "think" about the actions that should be done, have "hands" to interact with the digital world and even with the physical world and execute those same actions.

The potential of AI agents in public services

Public services are organized, to a large extent, around processes of a certain complexity such as the processing of aid and subsidies, the management of files and licenses or the citizen service itself through multiple channels. They are processes with many different steps, rules and actors, where repetitive tasks and manual work of reviewing documentation abound.

As can be seen in the  European Union's eGovernment Benchmark, eGovernment initiatives in recent decades have made it possible to move towards greater digitalisation of public services. However, the new wave of AI technologies, especially when foundational models are combined with agents, opens the door to a new leap to intelligently automate and orchestrate a large part of administrative processes.

In this context, autonomous agents would allow:

• Orchestrate end-to-end processes such as collecting data from different sources, proposing forms already completed, detecting inconsistencies in the documentation provided, or generating draft resolutions for validation by the responsible personnel.
• Act as "co-pilots" of public employees, preparing drafts, summaries or proposals for decisions that are then reviewed and validated, assisting in the search for relevant information or pointing out possible risks or incidents that require human attention.
• Optimise citizen service processes by  supporting tasks such as managing medical appointments, answering queries about the status of files, facilitating the payment of taxes or guiding people in choosing the most appropriate procedure for their situation.

Various analyses on AI in the public sector suggest that this type of intelligent automation, as in the private sector, can reduce waiting times, improve the quality of decisions and free up staff time for more value-added tasks. A recent report by PWC and Microsoft exploring the potential of Agent AI for the public sector sums up the idea well, noting that by incorporating Agent AI into public services, governments can improve responsiveness and increase citizen satisfaction, provided that the right safeguards are in place.

In addition, the implementation of autonomous agents allows us to dream of a transition from a reactive administration (which waits for the citizen to request a service) to a proactive administration that offers to do part of those same actions for us: from notifying us that a grant has been opened for which we probably meet the requirements,  to proposing the renewal of a license before it expires or reminding us of a medical appointment.

An illustrative example of the latter could be an AI agent that, based on data on available services and the information that the citizen himself has authorised to use, detects that a new aid has been published for actions to improve energy efficiency through the renovation of homes and sends a personalised notice to those who could meet the requirements. Even offering them a pre-filled draft application for review and acceptance. The final decision is still human, but the effort of seeking information, understanding conditions, and preparing documentation could be greatly reduced.

The role of open data

For an AI agent to be able to act in a useful and responsible way, they need to leverage on an environment rich in quality data and a robust data governance system. Among those assets needed to develop a good autonomous agent strategy, open data is important in at least three dimensions:

1. Fuel for decision-making: AI agents need information on current regulations, service catalogues, administrative procedures, socio-economic and demographic indicators, data on transport, environment, urban planning, etc. To this end, data quality and structure is of great importance as outdated, incomplete, or poorly documented data can lead agents to make costly mistakes. In the public sector, these mistakes can translate into unfair decisions that could ultimately lead to a loss of public trust.
2. Testbed for evaluating and auditing agents: Just as open data is important for evaluating generative AI models, it can also be important for testing and auditing autonomous agents. For example, simulating fictitious files with synthetic data based on real distributions to check how an agent acts in different scenarios. In this way, universities, civil society organizations and the administration itself can examine the behavior of agents and detect problems before scaling their use.
3. Transparency and explainability: Open data could help document where the data an agent uses came from, how it has been transformed, or which versions of the datasets were in place when a decision was made. This traceability contributes to explainability and accountability, especially when an AI agent intervenes in decisions that affect people's rights or their access to public services. If citizens can consult, for example, the criteria and data that are applied to grant aid, confidence in the system is reinforced.

The panorama of agent AI in Spain and the rest of the world

Although the concept of agent AI is recent, there are already initiatives underway in the public sector at an international level and they are also beginning to make their way in the European and Spanish context:

• The Government Technology Agency (GovTech) of Singapore has published an Agentic AI Primer guide  to guide developers and public officials on how to apply this technology, highlighting both its advantages and risks. In addition, the government is piloting the use of agents in various settings to reduce the administrative burden on social workers and support companies in complex licensing processes. All this in a controlled environment (sandbox) to test these solutions before scaling them.
• The UK government has published a specific note within its "AI Insights" documentation to explain what agent AI is and why it is relevant to government services. In addition, it has announced a tender to develop a "GOV.UK Agentic AI Companion" that will serve as an intelligent assistant for citizens from the government portal.
• The European Commission, within the framework of the Apply AI strategy and the GenAI4EU initiative, has launched calls to finance pilot projects that introduce scalable and replicable generative AI solutions in public administrations, fully integrated into their workflows. These calls seek precisely to accelerate the pace of digitalization through AI (including specialized agents) to improve decision-making, simplify procedures and make administration more accessible.

In Spain, although the label "agéntica AI" is not yet widely used, some experiences that go in that direction can already be identified. For example, different administrations are incorporating co-pilots based on generative AI to support public employees in tasks of searching for information, writing and summarizing documents, or managing files, as shown by initiatives of regional governments such as that of Aragon and local entities such as Barcelona City Council that are beginning to document themselves publicly.

The leap towards more autonomous agents in the public sector therefore seems to be a natural evolution on the basis of the existing e-government. But this evolution must, at the same time, reinforce the commitment to transparency, fairness, accountability, human oversight and regulatory compliance required by the AI Regulation and the rest of the regulatory framework and which should guide the actions of the public administration.

Looking to the Future: AI Agents, Open Data, and Citizen Trust

The arrival of agent AI once again offers the public administration new tools to reduce bureaucracy, personalize care and optimize its always scarce resources. However, technology is only a means, the ultimate goal is still  to generate public value by reinforcing the trust of citizens.

In principle, Spain is in a good position: it has an Artificial Intelligence Strategy 2024 that is committed to transparent, ethical and human-centred AI, with specific lines to promote its use in the public sector; it has aconsolidated open data infrastructure; and it has created the Spanish Agency for the Supervision of Artificial Intelligence (AESIA) as a body in charge of ensuring an ethical and safe use of AI, in accordance with the European AI Regulation.

We are, therefore, facing a new opportunity for modernisation that can build more efficient, closer and even proactive public services. If we are able to adopt the Agent AI properly, the agents that are deployed will not be a "black box" that acts without supervision, but digital, transparent and auditable "public agents", designed to work with open data, explain their decisions and leave a trace of the actions they take. Tools, in short, inclusive, people-centred and aligned with the values of public service.

Content created by Jose Luis Marín, Senior Consultant in Data, Strategy, Innovation & Digitalisation. The contents and views expressed in this publication are the sole responsibility of the author.

IA agents (such as Google ADK, Langchain and so) are so-called "brains". But these brains without "hands" cannot operate on the real world performing API requests or database queries. These "hands" are the tools. 

The challenge is the following: how do you connect brain with hands in an standard, decoupled and scalable fashion? The answers is the Model Context Protocol (MCP).

As a practical exercise, we built a conversational agent system that explores the Open Data national repository hosted at datos.gob.es through natural language questions, smoothing in this way the access to open data.  

In this practical exercise, the main objective is to illustrate, step by step, how to build an independent tools server that interacts with the MCP protocol. 

To make this exercise tangible and not just theoretical, we will use FastMCP to build the server. To prove that our server works, we will create a simple agent with Google ADK that uses it. The use case (querying the datos.gob.es API) illustrates this connection between tools and agents. The real learning lies in the architecture, which you could reuse for any API or database.

Below are the technologies we will use and a diagram showing how the different components are related to each other.

• FastMCP (mcp.server.fastmcp): a lightweight implementation of the MCP protocol that allows you to create tool servers with very little code using Python decorators. It is the “main character” of the exercise.

• Google ADK (Agent Development Kit): a framework to define the AI agent, its prompt, and connect it to the tools. It is the “client” that tests our server.

• FastAPI: used to serve the agent as a REST API with an interactive web interface.

• httpx: used to make asynchronous calls to the external datos.gob.es API.

• Docker and Docker Compose: used to package and orchestrate the two microservices, allowing them to run and communicate in isolation.

Figure 1. Decoupled architecture with MCP comunication.

Figure 1 illustrates a decoupled architecture divided into four main components that communicate via the MCP protocol. When the user makes a natural language query, the ADK Agent (based on Google Gemini) processes the intent and communicates with the MCP server through the MCP Protocol, which acts as a standardized intermediary. The MCP server exposes four specialized tools (search datasets, list topics, search by topic, and get details) that encapsulate all the business logic for interacting with the external datos.gob.es API. Once the tools execute the required queries and receive the data from the national catalog, the result is propagated back to the agent, which finally generates a user-friendly response, thus completing the communication cycle between the “brain” (agent) and the “hands” (tools).

Access the data lab repository on GitHub.

Run the data pre-processing code on Google Colab.

The architecture: MCP server and consumer agent

The key to this exercise is understanding the client–server relationship:

1. The Server (Backend): it is the protagonist of this exercise. Its only job is to define the business logic (the “tools”) and expose them to the outside world using the standard MCP “contract.” It is responsible for encapsulating all the logic for communicating with the datos.gob.es API.
2. The Agent (Frontend): it is the “client” or “consumer” of our server. Its role in this exercise is to prove that our MCP server works. We use it to connect, discover the tools that the server offers, and call them.
3. The MCP Protocol: it is the “language” or “contract” that allows the agent and the server to understand each other without needing to know the internal details of the other.

Development process

The core of the exercise is divided into three parts: creating the server, creating a client to test it, and running them.

1. The tool server (the backend with MCP)

This is where the business logic lives and the main focus of this tutorial. In the main file (server.py), we define simple Python functions and use the FastMCP @mcp.tool decorator to expose them as consumable “tools.”

The description we add to the decorator is crucial, since it is the documentation that any MCP client (including our ADK agent) will read to know when and how to use each tool.

The tools we will define in this exercise are:

• buscar_datasets(titulo: str): to search for datasets by keywords in the title.
• listar_tematicas(): to discover which data categories exist.
• buscar_por_tematica(tematica_id: str): to find datasets for a specific topic.
• obtener_detalle_dataset(dataset_id: str): to retrieve the complete information for a dataset.

2. The consumer agent (the frontend with Google ADK)

Once our MCP server is built, we need a way to test it. This is where Google ADK comes in. We use it to create a simple “consumer agent.”

The magic of the connection happens in the tools argument. Instead of defining the tools locally, we simply pass it the URL of our MCP server. When the agent starts, it will query that URL, read the MCP “contract,” and automatically know which tools are available and how to use them.

# Ejemplo de configuración en agent.py
root_agent = LlmAgent(
   ...
   instruction="Eres un asistente especializado en datos.gob.es...",
   tools=[
       MCPToolset(
           connection_params=StreamableHTTPConnectionParams(
               url="http://mcp-server:8000/mcp",
           ),
       )
   ]
)

3. Orchestration with Docker Compose

Finally, to run our MCP Server and the consumer agent together, we use docker-compose.yml. Docker Compose takes care of building the images for each service, creating a private network so they can communicate (which is why the agent can call http://mcp-server:8000), and exposing the necessary ports.

Testing the MCP server in action

Once we run docker-compose up --build, we can access the agent’s web interface at http://localhost:8080.

The goal of this test is not only to see whether the bot responds correctly, but to verify that our MCP server works properly and that the ADK agent (our test client) can discover and use the tools it exposes.

Figure 2. Screenshot of the agent showing its tools.

The true power of decoupling becomes evident when the agent logically chains together the tools provided by our server.

Figure 3. Screenshot of the agent showing the joint use of tools.

What can we learn?

The goal of this exercise is to learn the fundamentals of a modern agent architecture, focusing on the tool server. Specifically:

• How to build an MCP server: how to create a tool server from scratch that speaks MCP, using decorators such as @mcp.tool.
• The decoupled architecture pattern: the fundamental pattern of separating the “brain” (LLM) from the “tools” (business logic).
• Dynamic tool discovery: how an agent (in this case, an ADK agent) can dynamically connect to an MCP server to discover and use tools.
• External API integration: the process of “wrapping” a complex API (such as datos.gob.es) in simple functions within a tool server.
• Orchestration with Docker: how to manage a microservices project for development.

Conclusions and future work

We have built a robust and functional MCP tool server. The real value of this exercise lies in the how: a scalable architecture centered around a tool server that speaks a standard protocol.

This MCP-based architecture is incredibly flexible. The datos.gob.es use case is just one example. We could easily:

• Change the use case: replace server.py with one that connects to an internal database or the Spotify API, and any agent that speaks MCP (not just ADK) could use it.
• Change the “brain”: swap the ADK agent for a LangChain agent or any other MCP client, and our tool server would continue to work unchanged.

For those interested in taking this work to the next level, the possibilities focus on improving the MCP server:

• Implement more tools: add filters by format, publisher, or date to the MCP server.
• Integrate caching: use Redis in the MCP server to cache API responses and improve speed.
• Add persistence: store chat history in a database (this would be on the agent side).

Beyond these technical improvements, this architecture opens the door to many applications across very different contexts.

• Journalists and academics can have research assistants that help them discover relevant datasets in seconds.
• Transparency organizations can build monitoring tools that automatically detect new publications of public procurement or budget data.
• Consulting firms and business intelligence teams can develop systems that cross-reference information from multiple government sources to produce sector reports.
• Even in education, this architecture serves as a didactic foundation for teaching advanced concepts such as asynchronous programming, API integration, and AI agent design.

The pattern we have built—a decoupled tool server that speaks a standard protocol—is the foundation on which you can develop solutions tailored to your specific needs, regardless of the domain or data source you are working with.
 

Geomatico is a company specializing in the development of open source Geographic Information Systems (GIS). They offer customized web maps and GIS dashboards that add value to their clients' data.

Cities, infrastructures and the environment today generate a constant flow of data from sensors, transport networks, weather stations and Internet of Things (IoT) platforms, understood as networks of physical devices (digital traffic lights, air quality sensors, etc.) capable of measuring and transmitting information through digital systems. This growing volume of information makes it possible to improve the provision of public services, anticipate emergencies, plan the territory and respond to challenges associated with climate, mobility or resource management.

The increase in connected sources has transformed the nature of geospatial data. In contrast to traditional sets – updated periodically and oriented towards reference cartography or administrative inventories – dynamic data incorporate the temporal dimension as a structural component. An observation of air quality, a level of traffic occupancy or a hydrological measurement not only describes a phenomenon, but also places it at a specific time. The combination of space and time makes these observations fundamental elements for operating systems, predictive models and analyses based on time series.

In the field of open data, this type of information poses both opportunities and specific requirements. Opportunities include the possibility of building reusable digital services, facilitating near-real-time monitoring of urban and environmental phenomena, and fostering a reuse ecosystem based on continuous flows of interoperable data. The availability of up-to-date data also increases the capacity for evaluation and auditing of public policies, by allowing decisions to be contrasted with recent observations.

However, the opening of geospatial data in real time requires solving problems derived from technological heterogeneity. Sensor networks use different protocols, data models, and formats; the sources generate high volumes of observations with high frequency; and the absence of common semantic structures makes it difficult to cross-reference data between domains such as mobility, environment, energy or hydrology. In order for this data to be published and reused consistently, an interoperability framework is needed that standardizes the description of observed phenomena, the structure of time series, and access interfaces.

The open standards of the Open Geospatial Consortium (OGC) provide that framework. They define how to represent observations, dynamic entities, multitemporal coverages or sensor systems; establish APIs based on web principles that facilitate the consultation of open data; and allow different platforms to exchange information without the need for specific integrations. Its adoption reduces technological fragmentation, improves coherence between sources and favours the creation of public services based on up-to-date data.

Interoperability: The basic requirement for opening dynamic data

Public administrations today manage data generated by sensors of different types, heterogeneous platforms, different suppliers and systems that evolve independently. The publication of geospatial data in real time requires interoperability that allows information from multiple sources to be integrated, processed and reused. This diversity causes inconsistencies in formats, structures, vocabularies and protocols, which makes it difficult to open the data and reuse it by third parties. Let's see which aspects of interoperability are affected:

• Technical interoperability: refers to the ability of systems to exchange data using compatible interfaces, formats and models. In real-time data, this exchange requires mechanisms that allow for fast queries, frequent updates, and stable data structures. Without these elements, each flow would rely on ad hoc integrations, increasing complexity and reducing reusability.
• The Semantic interoperability: Dynamic data describe phenomena that change over short periods – traffic levels, weather parameters, flows, atmospheric emissions – and must be interpreted consistently. This implies having observation models, Vocabularies and common definitions that allow different applications to understand the meaning of each measurement and its units, capture conditions or constraints. Without this semantic layer, the opening of data in real time generates ambiguity and limits its integration with data from other domains.
• Structural interoperability: Real-time data streams tend to be continuous and voluminous, making it necessary to represent them as time series or sets of observations with consistent attributes. The absence of standardized structures complicates the publication of complete data, fragments information and prevents efficient queries. To provide open access to these data, it is necessary to adopt models that adequately represent the relationship between observed phenomenon, time of observation, associated geometry and measurement conditions.
• Interoperability in access via API: it is an essential condition for open data. APIs must be stable, documented, and based on public specifications that allow for reproducible queries. In the case of dynamic data, this layer guarantees that the flows can be consumed by external applications, analysis platforms, mapping tools or monitoring systems that operate in contexts other than the one that generates the data. Without interoperable APIs, real-time data is limited to internal uses.

Together, these levels of interoperability determine whether dynamic geospatial data can be published as open data without creating technical barriers.

OGC Standards for Publishing Real-Time Geospatial Data

The publication of georeferenced data in real time requires mechanisms that allow any user – administration, company, citizens or research community – to access them easily, with open formats and through stable interfaces. The Open Geospatial Consortium (OGC) develops a set of standards that enable exactly this: to describe, organize and expose spatial data in an interoperable and accessible way, which contributes to the openness of dynamic data.

What is OGC and why are its standards relevant?

The OGC is an international organization that defines common rules so that different systems can understand, exchange and use geospatial data without depending on specific technologies. These rules are published as open standards, which means that any person or institution can use them. In the realm of real-time data, these standards make it possible to:

• Represent what a sensor measures (e.g., temperature or traffic).
• Indicate where and when the observation was made.
• Structure time series.
• Expose data through open APIs.
• Connect IoT devices and networks with public platforms.

Together, this ecosystem of standards allows geospatial data – including data generated in real time – to be published and reused following a consistent framework. Each standard covers a specific part of the data cycle: from the definition of observations and sensors, to the way data is exposed using open APIs or web services. This modular organization makes it easier for administrations and organizations to select the components they need, avoiding technological dependencies and ensuring that data can be integrated between different platforms.

The OGC API family: Modern APIs for accessing open data

Within OGC, the newest line is family OGC API, a set of modern web interfaces designed to facilitate access to geospatial data using URLs and formats such as JSON or GeoJSON, common in the open data ecosystem.

Estas API permiten:

• Get only the part of the data that matters.
• Perform spatial searches ("give me only what's in this area").
• Access up-to-date data without the need for specialized software.
• Easily integrate them into web or mobile applications.

In this report: "How to use OGC APIs to boost geospatial data interoperability", we already told you about some of the most popular OGP APIs. While the report focuses on how to use OGC APIs for practical interoperability, this post expands the focus by explaining the underlying OGC data models—such as O&M, SensorML, or Moving Features—that underpin that interoperability.

On this basis, this post focuses on the standards that make this fluid exchange of information possible, especially in open data and real-time contexts. The most important standards in the context of real-time open data are:

Table 1. OGC Standards Relevant to Real-Time Geospatial Data

Models that structure observations and dynamic data

In addition to APIs, OGC defines several conceptual data models that allow you to consistently describe observations, sensors, and phenomena that change over time:

• O&M (Observations & Measurements): A model that defines the essential elements of an observation—measured phenomenon, instant, unity, and result—and serves as the semantic basis for sensor and time series data.
• SensorML: Language that describes the technical and operational characteristics of a sensor, including its location, calibration, and observation process.
• Moving Features: A model that allows mobile objects to be represented by means of space-time trajectories (such as vehicles, boats or fauna).

These models make it easy for different data sources to be interpreted uniformly and combined in analytics and applications.

The value of these standards for open data

Using OGC standards makes it easier to open dynamic data because:

• It provides common models that reduce heterogeneity between sources.
• It facilitates integration between domains (mobility, climate, hydrology).
• Avoid dependencies on proprietary technology.
• It allows the data to be reused in analytics, applications, or public services.
• Improves transparency by documenting sensors, methods, and frequencies.
• It ensures that data can be consumed directly by common tools.

Together, they form a conceptual and technical infrastructure that allows real-time geospatial data to be published as open data, without the need to develop system-specific solutions.

Real-time open geospatial data use cases

Real-time georeferenced data is already published as open data in different sectoral areas. These examples show how different administrations and bodies apply open standards and APIs to make dynamic data related to mobility, environment, hydrology and meteorology available to the public.

Below are several domains where Public Administrations already publish dynamic geospatial data using OGC standards.

Mobility and transport

Mobility systems generate data continuously: availability of shared vehicles, positions in near real-time, sensors for crossing in cycle lanes, traffic gauging or traffic light intersection status. These observations rely on distributed sensors and require data models capable of representing rapid variations in space and time.

OGC standards play a central role in this area. In particular, the OGC SensorThings API allows you to structure and publish observations from urban sensors using a uniform model – including devices, measurements, time series and relationships between them – accessible through an open API. This makes it easier for different operators and municipalities to publish mobility data in an interoperable way, reducing fragmentation between platforms.

The use of OGC standards in mobility not only guarantees technical compatibility, but also makes it possible for this data to be reused together with environmental, cartographic or climate information, generating multi-thematic analyses for urban planning, sustainability or operational transport management.

Example:

The open service of Toronto Bike Share, which publishes in SensorThings API format the status of its bike stations and vehicle availability.

Here each station is a sensor and each observation indicates the number of bicycles available at a specific time. This approach allows analysts, developers or researchers to integrate this data directly into urban mobility models, demand prediction systems or citizen dashboards without the need for specific adaptations.

Air quality, noise and urban sensors

Networks for monitoring air quality, noise or urban environmental conditions depend on automatic sensors that record measurements every few minutes. In order for this data to be integrated into analytics systems and published as open data, consistent models and APIs need to be available.

In this context, services based on OGC standards make it possible to publish data from fixed stations or distributed sensors in an interoperable way. Although many administrations use traditional interfaces such as OGC WMS to serve this data, the underlying structure is usually supported by observation models derived from the Observations & Measurements (O&M) family, which defines how to represent a measured phenomenon, its unit and the moment of observation.

Example:

The service Defra UK-AIR Sensor Observation Service provides access to near-real-time air quality measurement data from on-site stations in the UK.

The combination of O&M for data structure and open APIs for publication makes it easier for these urban sensors to be part of broader ecosystems that integrate mobility, meteorology or energy, enabling advanced urban analyses or environmental dashboards in near real-time.

Water cycle, hydrology and risk management

Hydrological systems generate crucial data for risk management: river levels and flows, rainfall, soil moisture or information from hydrometeorological stations. Interoperability is especially important in this domain, as this data is combined with hydraulic models, weather forecasting, and flood zone mapping.

To facilitate open access to time series and hydrological observations, several agencies use OGC API – Environmental Data Retrieval (EDR), an API designed to retrieve environmental data using simple queries at points, areas, or time intervals.

Example:

The USGS (United States Geological Survey), which documents the use of OGC API – EDR to access precipitation, temperature, or hydrological variable series.

This case shows how EDR allows you to request specific observations by location or date, returning only the values needed for analysis. While the USGS's specific hydrology data is served through its proprietary API, this case demonstrates how EDR fits into the hydrometeorological data structure and how it is applied in real operational flows.

The use of OGC standards in this area allows dynamic hydrological data to be integrated with flood zones, orthoimages or climate models, creating a solid basis for early warning systems, hydraulic planning and risk assessment.

Weather observation and forecasting

Meteorology is one of the domains with the highest production of dynamic data: automatic stations, radars, numerical prediction models, satellite observations and high-frequency atmospheric products. To publish this information as open data, the OGC API family  is becoming a key element, especially through OGC API – EDR, which allows observations or predictions to be retrieved in specific locations and at different time levels.

Example: 

The service NOAA OGC API – EDR, which provides access to weather data and atmospheric variables from the National Weather Service (United States).

This API allows data to be consulted at points, areas or trajectories, facilitating the integration of meteorological observations into external applications, models or services based on open data.

The use of OGC API in meteorology allows data from sensors, models, and satellites to be consumed through a unified interface, making it easy to reuse for forecasting, atmospheric analysis, decision support systems, and climate applications.

Best Practices for Publishing Open Geospatial Data in Real-Time 

The publication of dynamic geospatial data requires adopting practices that ensure its accessibility, interoperability, and sustainability. Unlike static data, real-time streams have additional requirements related to the quality of observations, API stability, and documentation of the update process. Here are some best practices for governments and organizations that manage this type of data.

• Stable open formats and APIs: The use of OGC standards – such as OGC API, SensorThings API or EDR – makes it easy for data to be consumed from multiple tools without the need for specific adaptations. APIs must be stable over time, offer well-defined versions, and avoid dependencies on proprietary technologies. For raster data or dynamic models, OGC services such as WMS, WMTS, or WCS are still suitable for visualization and programmatic access.
• DCAT-AP and OGC Models Compliant Metadata: Catalog interoperability requires describing datasets using profiles such as DCAT-AP, supplemented by O&M-based geospatial and observational metadata or SensorML. This metadata should document the nature of the sensor, the unit of measurement, the sampling rate, and possible limitations of the data.
• Quality, update frequency and traceability policies: dynamic datasets must explicitly indicate their update frequency, the origin of the observations, the validation mechanisms applied and the conditions under which they were generated. Traceability is essential for third parties to correctly interpret data, reproduce analyses and integrate observations from different sources.
• Documentation, usage limits, and service sustainability: Documentation should include usage examples, query parameters, response structure, and recommendations for managing data volume. It is important to set reasonable query limits to ensure the stability of the service and ensure that management can maintain the API over the long term.
• Licensing aspects for dynamic data: The license must be explicit and compatible with reuse, such as CC BY 4.0 or CC0. This allows dynamic data to be integrated into third-party services, mobile applications, predictive models or services of public interest without unnecessary restrictions. Consistency in the license also facilitates the cross-referencing of data from different sources.

These practices allow dynamic data to be published in a way that is reliable, accessible, and useful to the entire reuse community.

Dynamic geospatial data has become a structural piece for understanding urban, environmental and climatic phenomena. Its publication through open standards allows this information to be integrated into public services, technical analyses and reusable applications without the need for additional development. The convergence of observation models, OGC APIs, and best practices in metadata and licensing provides a stable framework for administrations and reusers to work with sensor data reliably. Consolidating this approach will allow progress towards a more coherent, connected public data ecosystem that is prepared for increasingly demanding uses in mobility, energy, risk management and territorial planning.

Content created by Mayte Toscano, Senior Consultant in Technologies related to the data economy. The content and views expressed in this publication are the sole responsibility of the author.

In the public sector ecosystem, subsidies represent one of the most important mechanisms for promoting projects, companies and activities of general interest. However, understanding how these funds are distributed, which agencies call for the largest grants or how the budget varies according to the region or beneficiaries is not trivial when working with hundreds of thousands of records.

In this line, we present a new practical exercise in the series "Step-by-step data exercises", in which we will learn how to explore and model open data using Apache Spark, one of the most widespread platforms for distributed processing and large-scale machine learning.

In this laboratory we will work with real data from the National System of Advertising of Subsidies and Public Aid (BDNS) and we will build a model capable of predicting the budget range of new calls based on their main characteristics.

All the code used is available in the corresponding GitHub repository so that you can run it, understand it, and adapt it to your own projects.

Access the datalab repository on GitHub

Run the data pre-processing code on Google Colab

Context: why analyze public subsidies?

The BDNS collects detailed information on hundreds of thousands of calls published by different Spanish administrations: from ministries and regional ministries to provincial councils and city councils. This dataset is an extraordinarily valuable source for:

• analyse the evolution of public spending,
• understand which organisms are most active in certain areas,
• identify patterns in the types of beneficiaries,
• and to study the budget distribution according to sector or territory.

In our case, we will use the dataset to address a very specific question, but of great practical interest:

Can we predict the budget range of a call based on its administrative characteristics?

This capability would facilitate initial classification, decision-making support or comparative analysis within a public administration.

Objective of the exercise

The objective of the laboratory is twofold:

1. Learn how to use Spark in a practical way:

• Upload a real high-volume dataset
• Perform transformations and cleaning
• Manipulate categorical and numeric columns
• Structuring a  machine learning pipeline

2. Building a predictive model

We will train a classifier capable of estimating whether a call belongs to one of these ranges of low budget (up to €20k), medium (between €20 and €150k) or high (greater than €150k), based on variables such as:

• Granting body
• Autonomous community
• Type of beneficiary
• Year of publication
• Administrative descriptions

Resources used

To complete this exercise we use:

Analytical tools

• Python, the main language of the project
• Google Colab, to run Spark and create Notebooks in a simple way
• PySpark, for data processing in the cleaning and modeling stages
• Pandas, for small auxiliary operations
• Plotly, for some interactive visualizations

Data

Official dataset of the National System of Advertising of Subsidies (BDNS), downloaded from the subsidy portal of the Ministry of Finance.

The data used in this exercise were downloaded on August 28, 2025. The reuse of data from the National System for the Publicity of Subsidies and Public Aid is subject to the legal conditions set out in https://www.infosubvenciones.es/bdnstrans/GE/es/avisolegal.

Development of the exercise

The project is divided into several phases, following the natural flow of a real Data Science case.

5.1. Data Dump and Transformation

In this first section we are going to automatically download the subsidy dataset from the API of the portal of the National System of Publicity of Subsidies (BDNS). We will then transform the data into an optimized format such as Parquet (columnar data format) to facilitate its exploration and analysis. 

In this process we will use some complex concepts, such as:

• Asynchronous functions: allows two or more independent operations to be processed in parallel, which makes it easier to make the process more efficient.
• Rotary writer: when a limit on the amount of information is exceeded, the file being processed is closed and a new one is opened with an auto-incremental index (after the previous one). This avoids processing files that are too large and improves efficiency.

Figure 1. Screenshot of the API of the National System for Advertising Subsidies and Public Aid

5.2. Exploratory analysis

The aim of this phase is to get a first idea of the characteristics of the data and its quality.

We will analyze, among others, aspects such as:

• Which types of subsidies have the highest number of calls.

Figure 2. Types of grants with the highest number of calls for applications.

• What is the distribution of subsidies according to their purpose (i.e. Culture, Education, Promotion of employment...).

Figure 3. Distribution of grants according to their purpose.

• Which purposes add a greater budget volume.

Figure 4. Purposes with the largest budgetary volume.

5.3. Modelling: construction of the budget classifier

At this point, we enter the most analytical part of the exercise: teaching a machine to predict whether a new call will have a low, medium or high budget based on its administrative characteristics. To achieve this, we designed a  complete machine learning pipeline in Spark that allows us to transform the data, train the model, and evaluate it in a uniform and reproducible way.

First, we prepare all the variables – many of them categorical, such as the convening body – so that the model can interpret them. We then combine all that information into a single vector that serves as the starting point for the learning phase.

With that foundation built, we train a classification model that learns to distinguish subtle patterns in the data: which agencies tend to publish larger calls or how specific administrative elements influence the size of a grant.

Once trained, we analyze their performance from different angles. We evaluate their ability to correctly classify the three budget ranges and analyze their behavior using metrics such as accuracy or the confusion matrix.

Figure 5. Accuracy metrics.

But we do not stop there: we also study which variables have had the greatest weight in the decisions of the model, which allows us to understand which factors seem most decisive when it comes to anticipating the budget of a call.

Figure 6. Variables that have had the greatest weight in the model's decisions.

Conclusions of the exercise

This laboratory will allow us to see how Spark simplifies the processing and modelling of high-volume data, especially useful in environments where administrations generate thousands of records per year, and to better understand the subsidy system after analysing some key aspects of the organisation of these calls.

Do you want to do the exercise?

If you're interested in learning more about using Spark and advanced public data analysis, you can access the repository and run the full Notebook step-by-step.

Content created by Juan Benavente, senior industrial engineer and expert in technologies related to the data economy. The content and views expressed in this publication are the sole responsibility of the author.

We live in an age where more and more phenomena in the physical world can be observed, measured, and analyzed in real time. The temperature of a crop, the air quality of a city, the state of a dam, the flow of traffic or the energy consumption of a building are no longer data that are occasionally reviewed: they are continuous flows of information that are generated second by second.

This revolution would not be possible without cyber-physical systems (CPS), a technology that integrates sensors, algorithms and actuators to connect the physical world with the digital world. But CPS does not only generate data: it can also be fed by open data, multiplying its usefulness and enabling evidence-based decisions.

In this article, we will explore what CPS is, how it generates massive data in real time, what challenges it poses to turn that data into useful public information, what principles are essential to ensure its quality and traceability, and what real-world examples demonstrate the potential for its reuse. We will close with a reflection on the impact of this combination on innovation, citizen science and the design of smarter public policies.

What are cyber-physical systems?

A cyber-physical system is a tight integration between digital components – such as software, algorithms, communication and storage – and physical components – sensors, actuators, IoT devices or industrial machines. Its main function is to observe the environment, process information and act on it.

Unlike traditional monitoring systems, a CPS is not limited to measuring: it closes a complete loop between perception, decision, and action. This cycle can be understood through three main elements:

Figure 1. Cyber-physical systems cycle. Source: own elaboration

An everyday example that illustrates this complete cycle of perception, decision and action very well is smart irrigation, which is increasingly present in precision agriculture and home gardening systems. In this case, sensors distributed throughout the terrain continuously measure soil moisture, ambient temperature, and even solar radiation. All this information flows to the computing unit, which analyzes the data, compares it with previously defined thresholds or with more complex models – for example, those that estimate the evaporation of water or the water needs of each type of plant – and determines whether irrigation is really necessary.

When the system concludes that the floor has reached a critical level of dryness, the third element of CPS comes into play: the actuators. They are the ones who open the valves, activate the water pump or regulate the flow rate, and they do so for the exact time necessary to return the humidity to optimal levels. If conditions change—if it starts raining, if the temperature drops, or if the soil recovers moisture faster than expected—the system itself adjusts its behavior accordingly.

This whole process happens without human intervention, autonomously. The result is a more sustainable use of water, better cared for plants and a real-time adaptability that is only possible thanks to the integration of sensors, algorithms and actuators characteristic of cyber-physical systems.

CPS as real-time data factories

One of the most relevant characteristics of cyber-physical systems is their ability to generate data continuously, massively and with a very high temporal resolution. This constant production can be seen in many day-to-day situations:

• A hydrological station can record level and flow every minute.
• An urban mobility sensor can generate hundreds of readings per second.
• A smart meter records electricity consumption every few minutes.
• An agricultural sensor measures humidity, salinity, and solar radiation several times a day.
• A mapping drone captures decimetric GPS positions in real time.

Beyond these specific examples, the important thing is to understand what this capability means for the system as a whole: CPS become true data factories, and in many cases come to function as digital twins of the physical environment they monitor. This almost instantaneous equivalence between the real state of a river, a crop, a road or an industrial machine and its digital representation allows us to have an extremely accurate and up-to-date portrait of the physical world, practically at the same time as the phenomena occur.

This wealth of data opens up a huge field of opportunity when published as open information. Data from CPS can drive innovative services developed by companies, fuel high-impact scientific research, empower citizen science initiatives that complement institutional data, and strengthen transparency and accountability in the management of public resources.

However, for all this value to really reach citizens and the reuse community, it is necessary to overcome a series of technical, organisational and quality challenges that determine the final usefulness of open data. Below, we look at what those challenges are and why they are so important in an ecosystem that is increasingly reliant on real-time generated information.

The challenge: from raw data to useful public information

Just because a CPS generates data does not mean that it can be published directly as open data. Before reaching the public and reuse companies, the information needs prior preparation , validation, filtering and documentation. Administrations must ensure that such data is understandable, interoperable and reliable. And along the way, several challenges appear.

One of the first is standardization. Each manufacturer, sensor and system can use different formats, different sample rates or its own structures. If these differences are not harmonized, what we obtain is a mosaic that is difficult to integrate. For data to be interoperable, common models, homogeneous units, coherent structures, and shared standards are needed. Regulations such as INSPIRE or the OGC (Open Geospatial Consortium) and IoT-TS standards are key so that data generated in one city can be understood, without additional transformation, in another administration or by any reuser.

The next big challenge is quality. Sensors can fail, freeze always reporting the same value, generate physically impossible readings, suffer electromagnetic interference or be poorly calibrated for weeks without anyone noticing. If this information is published as is, without a prior review and cleaning process, the open data loses value and can even lead to errors. Validation – with automatic checks and periodic review – is therefore indispensable.

Another critical point is contextualization. An isolated piece of information is meaningless. A "12.5" says nothing if we don't know if it's degrees, liters or decibels. A measurement of "125 ppm" is useless if we do not know what substance is being measured. Even something as seemingly objective as coordinates needs a specific frame of reference. And any environmental or physical data can only be properly interpreted if it is accompanied by the date, time, exact location and conditions of capture. This is all part of metadata, which is essential for third parties to be able to reuse information unambiguously.

It's also critical to address privacy and security. Some CPS can capture information that, directly or indirectly, could be linked to sensitive people, property, or infrastructure. Before publishing the data, it is necessary to apply anonymization processes, aggregation techniques, security controls and impact assessments that guarantee that the open data does not compromise rights or expose critical information.

Finally, there are operational challenges such as refresh rate and robustness of data flow. Although CPS generates information in real time, it is not always appropriate to publish it with the same granularity: sometimes it is necessary to aggregate it, validate temporal consistency or correct values before sharing it. Similarly, for data to be useful in technical analysis or in public services, it must arrive without prolonged interruptions or duplication, which requires a stable infrastructure and monitoring mechanisms.

Quality and traceability principles needed for reliable open data

Once these challenges have been overcome, the publication of data from cyber-physical systems must be based on a series of principles of quality and traceability. Without them, information loses value and, above all, loses trust.

The first is accuracy. The data must faithfully represent the phenomenon it measures. This requires properly calibrated sensors, regular checks, removal of clearly erroneous values, and checking that readings are within physically possible ranges. A sensor that reads 200°C at a weather station or a meter that records the same consumption for 48 hours are signs of a problem that needs to be detected before publication.

The second principle is completeness. A dataset should indicate when there are missing values, time gaps, or periods when a sensor has been disconnected. Hiding these gaps can lead to wrong conclusions, especially in scientific analyses or in predictive models that depend on the continuity of the time series.

The third key element is traceability, i.e. the ability to reconstruct the history of the data. Knowing which sensor generated it, where it is installed, what transformations it has undergone, when it was captured or if it went through a cleaning process allows us to evaluate its quality and reliability. Without traceability, trust erodes and data loses value as evidence.

Proper updating is another fundamental principle. The frequency with which information is published must be adapted to the phenomenon measured. Air pollution levels may need updates every few minutes; urban traffic, every second; hydrology, every minute or every hour depending on the type of station; and meteorological data, with variable frequencies. Posting too quickly can generate noise; too slow, it can render the data useless for certain uses.

The last principle is that of rich metadata. Metadata explains the data: what it measures, how it is measured, with what unit, how accurate the sensor is, what its operating range is, where it is located, what limitations the measurement has and what this information is generated for. They are not a footnote, but the piece that allows any reuser to understand the context and reliability of the dataset. With good documentation, reuse isn't just possible: it skyrockets.

Examples: CPS that reuses public data to be smarter

In addition to generating data, many cyber-physical systems also consume public data to improve their performance. This feedback makes open data a central resource for the functioning of smart territories. When a CPS integrates information from its own sensors with external open sources, its anticipation, efficiency, and accuracy capabilities are dramatically increased.

Precision agriculture: In agriculture, sensors installed in the field allow variables such as soil moisture, temperature or solar radiation to be measured. However, smart irrigation systems do not rely solely on this local information: they also incorporate weather forecasts from AEMET, open IGN maps  on slope or soil types, and climate models published as public data. By combining their own measurements with these external sources, agricultural CPS can determine much more accurately which areas of the land need water, when to plant, and how much moisture should be maintained in each crop. This fine management allows water and fertilizer savings that, in some cases, exceed 30%.

Water management: Something similar happens in water management. A cyber-physical system that controls a dam or irrigation canal needs to know not only what is happening at that moment, but also what may happen in the coming hours or days. For this reason, it integrates its own level sensors with open data on river gauging, rain and snow predictions, and even public information on ecological flows. With this expanded vision, the CPS can anticipate floods, optimize the release of the reservoir, respond better to extreme events or plan irrigation sustainably. In practice, the combination of proprietary and open data translates into safer and more efficient water management.

Impact: innovation, citizen science, and data-driven decisions

The union between cyber-physical systems and open data generates a multiplier effect that is manifested in different areas.

• Business innovation: Companies have fertile ground to develop solutions based on reliable and real-time information. From open data and CPS measurements, smarter mobility applications, water management platforms, energy analysis tools, or predictive systems for agriculture can emerge. Access to public data lowers barriers to entry and allows services to be created without the need for expensive  private datasets, accelerating innovation and the emergence of new business models.
• Citizen science: the combination of SCP and open data also strengthens social participation. Neighbourhood communities, associations or environmental groups can deploy low-cost sensors to complement public data and better understand what is happening in their environment. This gives rise to initiatives that measure noise in school zones, monitor pollution levels in specific neighbourhoods, follow the evolution of biodiversity or build collaborative maps that enrich official information.
• Better public decision-making: finally, public managers benefit from this strengthened data ecosystem. The availability of reliable and up-to-date measurements makes it possible to design low-emission zones, plan urban transport more effectively, optimise irrigation networks, manage drought or flood situations or regulate energy policies based on real indicators. Without open data that complements and contextualizes the information generated by the CPS, these decisions would be less transparent and, above all, less defensible to the public.

In short, cyber-physical systems have become an essential piece of understanding and managing the world around us. Thanks to them, we can measure phenomena in real time, anticipate changes and act in a precise and automated way. But its true potential unfolds when its data is integrated into a quality open data ecosystem, capable of providing context, enriching decisions and multiplying uses.

The combination of SPC and open data allows us to move towards smarter territories, more efficient public services and more informed citizen participation. It provides economic value, drives innovation, facilitates research and improves decision-making in areas as diverse as mobility, water, energy or agriculture.

For all this to be possible, it is essential to guarantee the quality, traceability and standardization of the published data, as well as to protect privacy and ensure the robustness of information flows. When these foundations are well established, CPS not only measure the world: they help it improve, becoming a solid bridge between physical reality and shared knowledge.

Content created by Dr. Fernando Gualo, Professor at UCLM and Government and Data Quality Consultant. The content and views expressed in this publication are the sole responsibility of the author.

Quantum computing promises to solve problems in hours that would take millennia for the world's most powerful supercomputers. From designing new drugs to optimizing more sustainable energy grids, this technology will radically transform our ability to address humanity's most complex challenges. However, its true democratizing potential will only be realized through convergence with open data, allowing researchers, companies, and governments around the world to access both quantum computing power in the cloud and the  public datasets needed to train and validate quantum algorithms.

Trying to explain quantum theory has always been a challenge, even for the most brilliant minds humanity has given in the last 2 centuries. The celebrated physicist Richard Feynman (1918-1988) put it with his trademark humor:

"There was a time when newspapers said that only twelve men understood the theory of relativity. I don't think it was ever like that [...] On the other hand, I think I can safely say that no one understands quantum mechanics." 

And that was said by one of the most brilliant physicists of the twentieth century, Nobel Prize winner and one of the fathers of quantum electrodynamics. So great is the rarity of quantum behavior in the eyes of a human that, even Albert Einstein himself, in his now mythical phrase, said to Max Born, in a letter written to the German physicist in 1926, "God does not play dice with the universe" in reference to his disbelief about the probabilistic and non-deterministic properties attributed to quantum behavior. To which Niels Bohr - another titan of physics of the twentieth century - replied: "Einstein, stop telling God what to do."

Classical computing

If we want to understand why quantum mechanics proposes a revolution in computer science, we have to understand its fundamental differences from mechanics - and, therefore, - classical computing. Almost all of us have heard of bits of information at some point in our lives. Humans have developed a way of performing complex mathematical calculations by reducing all information to bits - the fundamental units of information with which a machine knows how to work -, which are the famous zeros and ones (0 and 1). With two simple values, we have been able to model our entire mathematical world. And why? Some will ask. Why base 2 and not 5 or 7? Well, in our classic physical world (in which we live day to day) differentiating between 0 and 1 is relatively simple; on and off, as in the case of an electrical switch, or north or south magnetization, in the case of a magnetic hard drive. For a binary world, we have developed an entire coding language based on two states: 0 and 1.

Quantum computing

In quantum computing, instead of bits, we use qubits. Qubits use several "strange" properties of quantum mechanics that allow them to represent infinite states at once between zero and one of the classic bits. To understand it, it's as if a bit could only represent an on or off state in a light bulb, while a qubit can represent all the light bulb's illumination intensities. This property is known as "quantum superposition" and allows a quantum computer to explore millions of possible solutions at the same time. But this is not all in quantum computing. If quantum superposition seems strange to you, wait until you see quantum entanglement. Thanks to this property, two  "entangled" particles (or two qubits) are connected "at a distance" so that the state of one determines the state of the other. So, with these two properties we have information qubits, which can represent infinite states and are connected to each other. This system potentially has an exponentially greater computing capacity than our computers based on classical computing.     

Two application cases of quantum computing

1. Drug discovery and personalized medicine. Quantum computers can simulate complex molecular interactions that are impossible to compute with classical computing. For example, protein folding – fundamental to understanding diseases such as Alzheimer's – requires analyzing trillions of possible configurations. A quantum computer could shave years of research to weeks, speeding up the development of new drugs and personalized treatments based on each patient's genetic profile.

2. Logistics optimization and climate change. Companies like Volkswagen already use quantum computing to optimize traffic routes in real time. On a larger scale, these systems could revolutionize the energy management of entire cities, optimizing smart grids that integrate renewables efficiently, or design new materials for CO₂ capture that help combat climate change.

A good read recommended for a complete review of quantum computing here.

The role of open data (and computing resources)

The democratization of access to quantum computing will depend crucially on two pillars: open computing resources and  quality public datasets. This combination is creating an ecosystem where quantum innovation no longer requires millions of dollars in infrastructure. Here are some options available for each of these pillars.

1. Free access to real quantum hardware:

• IBM Quantum Platform: Provides free monthly access to quantum systems of more than 100 qubits for anyone in the world. With more than 400,000 registered users who have generated more than 2,800 scientific publications, it demonstrates how open access accelerates research. Any researcher can sign up for the platform and start experimenting in minutes.
• Open Quantum Institute (OQI): launched at CERN (the European Organization for Nuclear Research) in 2024, it goes further, providing not only access to quantum computing but also mentoring and educational resources for underserved regions. Its hackathon program in 2025 includes events in Lebanon, the United Arab Emirates, and other countries, specifically designed to mitigate the quantum digital divide.

2. Public datasets for the development of quantum algorithms:

• QDataSet: Offers 52 public datasets with simulations of one- and two-qubit quantum systems, freely available for training  quantum machine learning (ML) algorithms. Researchers without resources to generate their own simulation data can access its repository on GitHub and start developing algorithms immediately.
• ClimSim: This is a public climate-related modeling dataset that is already being used to demonstrate the first quantum ML algorithms applied to climate change. It allows any team, regardless of their budget, to work on real climate problems using quantum computing.
• PennyLane Datasets: is  an open collection of molecules, quantum circuits, and physical systems that allows  pharmaceutical startups without resources to perform expensive simulations and experiment with quantum-assisted drug discovery.

Real cases of inclusive innovation

The possibilities offered by the use of open data to quantum computing have been evident in various use cases, the result of specific research and calls for grants, such as:

• The Government of Canada launched in 2022 "Quantum Computing for Climate", a specific call for SMEs and startups to develop quantum applications using public climate data, demonstrating how governments can catalyze innovation by providing both data and financing for its use.
• The UK Quantum Catalyst Fund (£15 million) funds projects that combine quantum computing with public data from the UK's National Health Service (NHS) for problems such as optimising energy grids and medical diagnostics, creating solutions of public interest verifiable by the scientific community.
• The  Open Quantum Institute's (OQI) 2024 report details 10 use cases for the UN Sustainable Development Goals developed collaboratively by experts from 22 countries, where the results and methodologies are publicly accessible, allowing any institution to replicate or improve these works).
   
• Red.es has opened an expression of interest aimed at agents in the quantum technologies ecosystem to collect ideas, proposals and needs that contribute to the design of the future lines of action of the National Strategy for Quantum Technologies 2025–2030, financed with 40 million euros from the ERDF Funds.

Current State of Quantum Computing

We are in the NISQ (Noisy Intermediate-Scale Quantum) era, a term coined by physicist John Preskill in 2018, which describes quantum computers with 50-100  physical qubits. These systems are powerful enough to perform certain calculations beyond the classical capabilities, but they suffer from incoherence, frequent errors that make them unviable in market applications.

IBM, Google, and startups like IonQ offer cloud access to their quantum systems, with IBM providing public access through the IBM Quantum Platform since 2016, being one of the first publicly accessible quantum processors connected to the cloud.

In 2019, Google achieved "quantum supremacy" with its 53-qubit Sycamore processor, which performed a calculation in about 200 seconds that would take about 10,000 years to a state-of-the-art classical supercomputer.

The latest independent analyses suggest that practical quantum applications may emerge around 2035-2040, assuming continued exponential growth in quantum hardware capabilities. IBM has committed to delivering a large-scale fault-tolerant quantum computer, IBM Quantum Starling, by 2029, with the goal of running quantum circuits comprising 100 million quantum gates on 200  logical qubits.

The Global Race for Quantum Leadership

International competition for dominance in quantum technologies has triggered an unprecedented wave of investment. According to McKinsey, until 2022 the officially recognized level of public investment in China (15,300 million dollars) exceeds that of the European Union (7,200 million dollars), the United States 1,900 million dollars) and Japan (1,800 million dollars) combined.

Domestically, the UK has committed £2.5 billion over ten years to its National Quantum Strategy to make the country a global hub for quantum computing, and Germany has made one of the largest strategic investments in quantum computing, allocating €3 billion under its economic stimulus plan.

Investment in the first quarter of 2025 shows explosive growth: quantum computing companies raised more than $1.25 billion, more than double the previous year, an increase of 128%, reflecting a growing confidence that this technology is approaching commercial relevance.

To end the section, a fantastic short interview with Ignacio Cirac, one of the "Spanish fathers" of quantum computing.

Quantum Spain Initiative

In the case of Spain, 60 million euros have been invested in Quantum Spain, coordinated by the Barcelona Supercomputing Center. The project includes:

• Installation of the first quantum computer in southern Europe.
• Network of 25 research nodes distributed throughout the country.
• Training of quantum talent in Spanish universities.
• Collaboration with the business sector for real-world use cases.

This initiative positions Spain as a quantum hub in southern Europe, crucial for not being technologically dependent on other powers.

In addition, Spain's Quantum Technologies Strategy has recently been presented with an investment of 800 million euros. This strategy is structured into 4 strategic objectives and 7 priority actions.

Strategic objectives:

• Strengthen R+D+I to promote the transfer of knowledge and facilitate research reaching the market.
• To create a Spanish quantum market, promoting the growth and emergence of quantum companies and their ability to access capital and meet demand.
• Prepare society for disruptive change, promoting security and reflection on a new digital right, post-quantum privacy.
• Consolidate the quantum ecosystem in a way that drives a vision of the country.

Priority actions:

• Priority 1: To promote Spanish companies in quantum technologies.
• Priority 2: Develop algorithms and technological convergence between AI and Quantum.
• Priority 3: Position Spain as a benchmark in quantum communications.
• Priority 4: Demonstrate the impact of quantum sensing and metrology.
• Priority 5: Ensure the privacy and confidentiality of information in the post-quantum world.
• Priority 6: Strengthening capacities: infrastructure, research and talent.
• Priority 7: Develop a solid, coordinated and leading Spanish quantum ecosystem in the EU.

Figure 1. Spain's quantum technology strategy. Source: Author's own elaboration

In short, quantum computing and open data represent a major technological evolution that affects the way we generate and apply knowledge. If we can build a truly inclusive ecosystem—where access to quantum hardware, public datasets, and specialized training is within anyone's reach—we will open the door to a new era of collaborative innovation with a major global impact.

Content created by Alejandro Alija, expert in Digital Transformation and Innovation. The content and views expressed in this publication are the sole responsibility of the author.

The convergence between open data, artificial intelligence and environmental sustainability poses one of the main challenges for the digital transformation model that is being promoted at European level. This interaction is mainly materialized in three outstanding manifestations:

• The opening of high-value data directly related to sustainability, which can help the development of artificial intelligence solutions aimed at climate change mitigation and resource efficiency.

• The promotion of the so-called green algorithms in the reduction of the environmental impact of AI, which must be materialized both in the efficient use of digital infrastructure and in sustainable decision-making.

• The commitment to environmental data spaces, generating digital ecosystems where data from different sources is shared to facilitate the development of interoperable projects and solutions with a relevant impact from an environmental perspective.

Below, we will delve into each of these points.

High-value data for sustainability

 Directive (EU) 2019/1024 on open data and re-use of public sector information introduced for the first time the concept of high-value datasets, defined as those with exceptional potential to generate social, economic and environmental benefits. These sets should be published free of charge, in machine-readable formats, using application programming interfaces (APIs) and, where appropriate, be available for bulk download. A number of priority categories have been identified for this purpose, including environmental and Earth observation data.

This is a particularly relevant category, as it covers both data on climate, ecosystems or environmental quality, as well as those linked to the INSPIRE Directive, which refer to certainly diverse areas such as hydrography, protected sites, energy resources, land use, mineral resources or, among others, those related to areas of natural hazards, including orthoimages.

These data are particularly relevant when it comes to monitoring variables related to climate change, such as land use, biodiversity management taking into account the distribution of species, habitats and protected sites, monitoring of invasive species or the assessment of natural risks. Data on air quality and pollution are crucial for public and environmental health, so access to them allows  exhaustive analyses to be carried out,  which are undoubtedly relevant for the adoption of public policies aimed at improving them. The management of water resources can also be optimized through hydrography data and environmental monitoring, so that its massive and automated treatment is an inexcusable premise to face the challenge of the digitalization of water cycle management.

Combining it with other quality environmental data facilitates the development of AI solutions geared towards specific climate challenges. Specifically, they allow predictive models to be trained to anticipate extreme phenomena (heat waves, droughts, floods), optimize the management of natural resources or monitor critical environmental indicators in real time. It also makes it possible to promote high-impact economic projects, such as the use of AI algorithms to implement technological solutions in the field of precision agriculture, enabling the intelligent adjustment of irrigation systems, the early detection of pests or the optimization of the use of fertilizers.

Green algorithms and digital responsibility: towards sustainable AI

Training and deploying AI systems, particularly general-purpose models and large language models, involves significant energy consumption. According to estimates by the International Energy Agency, data centers accounted for around 1.5% of global electricity consumption in 2024. This represents a growth of around 12% per year since 2017, more than four times faster than the rate of total electricity consumption. Data center power consumption is expected to double to around 945 TWh by 2030.

Against this backdrop, green algorithms are an alternative that must necessarily be taken into account when it comes to minimising the environmental impact posed by the implementation of digital technology and, specifically, AI. In fact, both the European Data Strategy and the European Green Deal explicitly integrate digital sustainability as a strategic pillar. For its part, Spain has launched a National Green Algorithm Programme, framed in the 2026 Digital Agenda and with a specific measure in the National Artificial Intelligence Strategy.

One of the main objectives of the Programme is to promote the development of algorithms that minimise their environmental impact from conception ( green by design), so the requirement of exhaustive documentation of the datasets used to train AI models – including origin, processing, conditions of use and environmental footprint – is essential to fulfil this aspiration. In this regard, the Commission has published a template to help general-purpose AI providers summarise the data used for the training of their models, so that greater transparency can be demanded, which, for the purposes of the present case, would also facilitate traceability and responsible governance from an environmental perspective.  as well as the performance of eco-audits.

The European Green Deal Data Space

It is one of the common European data spaces contemplated in the European Data Strategy that is at a more advanced stage, as demonstrated by the numerous initiatives and dissemination events that have been promoted around it. Traditionally, access to environmental information has been one of the areas with the most favourable regulation, so that with the promotion of high-value data and the firm commitment to the creation of a European area in this area, there has been a very remarkable qualitative advance that reinforces an already consolidated trend in this area.

Specifically, the data spaces model facilitates interoperability between public and private open data, reducing barriers to entry for startups and SMEs in sectors such as smart forest management, precision agriculture or, among many other examples, energy optimization. At the same time, it reinforces the quality of the data available for Public Administrations to carry out their public policies, since their own sources can be contrasted and compared with other data sets. Finally, shared access to data and AI tools can foster collaborative innovation initiatives and projects, accelerating the development of interoperable and scalable solutions.

However, the legal ecosystem of data spaces entails a complexity inherent in its own institutional configuration, since it brings together several subjects and, therefore, various interests and applicable legal regimes: 

• On the one hand, public entities, which have a particularly reinforced leadership role in this area.

• On the other hand, private entities and citizens, who can not only contribute their own datasets, but also offer digital developments and tools that value data through innovative services.

• And, finally, the providers of the infrastructure necessary for interaction within the space.

Consequently,  advanced governance models are essential to deal with this complexity, reinforced by technological innovation and especially AI, since the traditional approaches of legislation regulating access to environmental information are certainly limited for this purpose.

Towards strategic convergence

The convergence of high-value open data, responsible green algorithms and environmental data spaces is shaping a new digital paradigm that is essential to address climate and ecological challenges in Europe that requires a robust and, at the same time, flexible legal approach. This unique ecosystem not only allows innovation and efficiency to be promoted in key sectors such as precision agriculture or energy management, but also reinforces the transparency and quality of the environmental information available for the formulation of more effective public policies.

Beyond the current regulatory framework, it is essential to design governance models that help to interpret and apply diverse legal regimes in a coherent manner, that protect data sovereignty and, ultimately, guarantee transparency and responsibility in the access and reuse of environmental information. From the perspective of sustainable public procurement, it is essential to promote procurement processes by public entities that prioritise technological solutions and interoperable services based on open data and green algorithms, encouraging the choice of suppliers committed to environmental responsibility and transparency in the carbon footprints of their digital products and services.

Only on the basis of this approach can we aspire to make digital innovation technologically advanced and environmentally sustainable, thus aligning the objectives of the Green Deal, the European Data Strategy and the European approach to AI.

Content prepared by Julián Valero, professor at the University of Murcia and coordinator of the Innovation, Law and Technology Research Group (iDerTec). The content and views expressed in this publication are the sole responsibility of the author.