A city is never in “test mode”
Traffic flows shift throughout the day, public transport demand changes in real time, digital services increasingly support interactions between citizens and public administrations. In European cities, complexity, far from being an exception, is the operating condition.
Artificial Intelligence is now expected to function within this environment: not in controlled settings, but in dynamic, interconnected systems where technologies must adapt continuously and remain aligned with public governance frameworks.
This is what makes AI in urban contexts fundamentally different, because cities cannot be paused or simplified, and solutions must operate in real time, interacting with human behaviour and fitting within existing institutional and infrastructural constraints.
Operating as a Testing and Experimentation Facility for smart cities, CitCom.ai enables the validation of AI solutions in real urban environments, aligning technological innovation with the frameworks that underpin urban systems.
Public sector innovation as a policy space
Many AI solutions for smart cities begin as small-scale pilots, often disconnected from the complexity of real urban systems.
Yet the transition to deployment introduces structural challenges: fragmented governance, heterogeneous infrastructures and strict public procurement and accountability requirements.
CitCom.ai addresses this gap by validating solutions in real or highly realistic urban environments. These include applications in mobility, energy, environmental monitoring and digital public services. In all cases, technologies are assessed not only for performance, but for their ability to integrate into complex urban systems and operate reliably over time.
The focus shifts from isolated experimentation to system-level validation, in a context in which AI in cities is not only a technological issue, but also a governance matter.
Local authorities play a dual role as both users and regulators of innovation, responsible for ensuring transparency, accountability and continuity of public services.
Testing and experimentation thus become tools for both technical validation and policy learning, enabling public authorities to better understand how AI behaves in real operational contexts.
At the same time, the initiative helps lower barriers for SMEs and innovators engaging with the public sector.
Testing and urban impact
The value of CitCom.ai becomes most visible when AI is embedded in real urban use cases.
A clear example comes from the success story of Mechelen, a Flemish city that used CitCom.ai to address a concrete challenge: improving cycling safety and infrastructure quality.
Through an innovation challenge structured as a public innovation procurement, the city invited companies to test solutions in real conditions. Among the selected participants was XenomatiX, a Belgian scale-up specialising in high-precision mapping using lidar and camera technology.
Within this framework, XenomatiX deployed its XenoBike system to map more than 50 kilometres of cycling infrastructure, focusing on key urban and inter-municipal routes.
The result is a detailed, high-resolution dataset that enables the city to identify priority areas for maintenance and improvement, turning spatial data into actionable planning insights.
Beyond the technical outcome, the case illustrates a broader shift: cities moving from passive recipients of technology to active environments where innovation is tested, shaped and directly linked to public value creation.
Building on this, CitCom.ai is part of a wider European effort to build a coherent ecosystem for AI in cities and communities.
By connecting municipalities, research organisations and technology providers across countries, it supports collaboration and contributes to the emergence of shared standards and interoperable solutions.
This is particularly important in a European context marked by diversity in governance models, infrastructure maturity and digital capacity.
To learn more about CitCom.ai, its services and activities, please visit: https://citcomtef.eu/.
- Logga in för att kommentera
Kommentarer
Som svar på First of all,… av remy wehrung
Thank you, Remy, for taking the time to share such a thoughtful and detailed perspective.
We fully agree with one of your central points: cities are complex systems where technology never operates in isolation. AI solutions interact with infrastructure, governance processes, public services, and, most importantly, citizens. In our experience, successful urban AI adoption depends as much on understanding these interactions and stakeholder perspectives as it does on the technology itself.
Many of the questions you raise around topics such as edge AI, cybersecurity, compliance, and trustworthiness are indeed critical considerations. Fortunately, these are also areas in which CitCom.ai brings together expertise from research institutes, cities, and technology partners across Europe. For readers interested in these aspects, I would encourage them to have a look at our service catalogue: https://citcomtef.eu/services/. My own organization imecfor example, has a strong focus on edge AI, which we see as an important building block for more resilient, sovereign, and privacy-conscious smart city solutions.
In fact, we see CitCom.ai as precisely the kind of instrument needed to address the questions you raise. What makes the initiative unique is the possibility to combine real-world experimentation with multidisciplinary expertise. Through TEFs such as citcom.ai, AI solutions can be assessed not only for performance, but also from the perspectives of safety, resilience, compliance, trustworthiness, and societal fit.
The goal is not to deploy or test AI at all costs, but to create the conditions for evaluating it critically. Often, the most valuable outcome of testing is a better understanding of whether a solution is fit for purpose, which risks need to be mitigated, and under which conditions it can be adopted responsibly.
Thank you again for contributing to the discussion. Comments like yours help highlight why testing and validating AI in real urban environments is so important in the first place.
First of all, congratulations on this initiative. Framing Artificial Intelligence as a concrete municipal policy instrument, rather than as an abstract technological narrative, is probably one of the most effective ways to make this family of technologies understandable and acceptable to citizens.
By linking AI experimentation to tangible urban outcomes such as mobility, infrastructure quality, environmental monitoring and public services, CitCom.ai contributes to repositioning AI as an operational public-interest capability rather than a purely commercial or theoretical field.
That being said, your presentation also raises several important — and ultimately very legitimate — questions regarding operational architecture, resilience, governance and public trust.
The first concerns infrastructure dependency.
Your framework appears to rely heavily on large-scale computational environments and potentially centralized datacenter infrastructures. In the context of real-time urban systems, why not privilege a stronger edge AI approach?
Cities are latency-sensitive environments where transport systems, environmental sensors, emergency coordination and public services increasingly require localised decision-making capacities. Edge architectures could potentially reduce dependency on centralized infrastructures, improve operational continuity during network disruption, and reinforce strategic digital resilience at municipal level.
This naturally leads to another key issue: operational continuity.
You correctly emphasize that cities evolve constantly and cannot be paused. But how does CitCom.ai itself maintain continuous operational capability in such conditions?
What mechanisms are implemented to ensure resilience against:
The document also mentions governance and accountability, but says relatively little regarding cybersecurity and citizen data protection. Yet in smart city environments, AI systems may indirectly interact with highly sensitive behavioural, mobility or administrative datasets.
Could you clarify:
Similarly, while alignment with the AI Act is implicitly suggested through your governance-oriented positioning, it would be valuable to understand more explicitly:
Another important dimension concerns testing methodology itself.
The Mechelen example is particularly interesting because it demonstrates real-world deployment beyond laboratory conditions. However, the text does not specify the human dimension of experimentation.
How were the validation protocols designed?
What population samples or representative user groups were involved?
Were citizens included directly in evaluation loops?
How were social acceptability, behavioural adaptation and accessibility measured alongside technical performance?
This question is essential because urban AI systems do not operate in purely technical ecosystems; they operate within human communities shaped by trust, diversity, social asymmetries and institutional expectations.
Overall, CitCom.ai appears to represent a highly relevant European initiative precisely because it approaches AI not merely as software deployment, but as an infrastructural and governance challenge embedded within democratic urban systems.
The next major step may therefore not simply concern technological scaling, but the capacity to demonstrate long-term resilience, security, transparency and public legitimacy under real operational conditions.