An operating system for smart cities

A Cologne-based startup has created what it claims is the world’s first operating system for smart cities.

The municipal operating system, ‘urbanOS’, was developed by dataMatters, a spin-off of RWTH Aachen University. It functions similarly to computer or smartphone operating systems, but is optimised for the ‘smartification’ of urban infrastructure, ranging from traffic management and waste disposal to energy supply and public safety.

According to the company, this ‘digital town hall’ is highly scalable — capable of meeting the needs of cities, districts and municipalities, from small towns to large metropolitan areas.

“The term ‘smart city’ describes a comprehensive concept for a city where data, in the form of smart — ‘intelligent’ — infrastructures, plays a key role. This includes the following areas: buildings, mobility, energy, water, waste management, health care and digital infrastructures,” explained Dr Daniel Trauth, founder and CEO of dataMatters GmbH.

“All these areas are covered by urbanOS.”

Planning ahead

The new municipal operating system is based on a multi-layered model: sensors, municipal data space, artificial intelligence (AI), data-driven decision-making, and optimised services for both citizens and local governments.

“A wide range of sensors captures what’s happening in the city,” Trauth said.

“This information is transferred to a municipal data space, where it is processed using AI. The results are displayed in an urbanCockpit, allowing decision-makers to gain a real-time overview of what’s actually happening in their city, similar to a pilot steering an aircraft.”

Thanks to the AI’s predictive capabilities, the urbanCockpit also provides a detailed forecast of future conditions. “This gives municipalities an unprecedented factual basis for urban decision-making,” Trauth said, giving the example of cameras in buses and trains, which “can precisely capture how many seats and standing spots are occupied at different times on various lines”.

“Based on this data, the AI can generate recommendations to optimise public transportation, factoring in events like festivals, football games or cultural events. The result is greater public transportation acceptance among citizens, more targeted staffing, and reductions in costs and environmental impact.”

Protecting sensitive data

According to dataMatters, urbanOS is the world’s first operating system to integrate federated AI (federated learning) from the outset. This means that an AI model is trained across numerous devices without centralising sensitive data.

The operating system also relies on edge computing, so that all data is cleaned of any personal references at the moment of collection. For example, cameras that count pedestrians, cyclists and vehicles to optimise traffic flow delete all faces and licence plates on-site. No information is transmitted to the municipal data space that could directly or indirectly lead to identifying individuals. The company describes this as “data anonymisation at the point of collection”.

“Municipalities want a smart city, but not a surveillance city,” Trauth explained. “Therefore, protecting citizens’ sensitive data is of the utmost priority.”

For urbanOS, all data is processed exclusively in data centres located in Germany and subject only to German law, the company said. No data is transferred to foreign data clouds — an important consideration given the patchy data protection in countries like the USA, where many common cloud providers are based.

“The much-demanded autonomy in digitalisation is fully ensured at the municipal level with urbanOS,” Trauth said.

Flexible AI

The type of artificial intelligence used within urbanOS can be freely chosen by each municipality or city. “We can integrate any large language model,” Trauth said.

It is also up to the municipal authorities whether the AI should be operated in a local data centre or sourced externally through a cloud service. The same applies to the data space where all municipal information converges to be ready for AI analysis. “If needed, we provide both the data space and artificial intelligence as a complete package,” Trauth said. “In this case, the municipality receives the full functionality, including AI analysis, from us, just like from a smart power plant.” Alternatively, the municipality can choose to operate the data space and AI system independently and only use dataMatters’ operating system for ‘smartification’.

According to dataMatters, urbanOS already supports all common AI models and IT systems, allowing them to be seamlessly integrated into the municipal operating system. Technically, this is achieved through an application programming interface (API) that encompasses all standard market interfaces. “We support countless connectors to sensors, wireless networks, database systems, as well as third-party software and OpenData portals,” Trauth said.

For the wireless connectivity of sensors, dataMatters relies on long range wide area networks (LoRaWAN). These networks enable data transmission over long distances with minimal energy consumption. The LoRa radio standard was specifically developed for the Internet of Things (IoT), which includes the various sensors used in smart cities. It is estimated that well over 10 billion IoT devices are currently in use worldwide.

Real-world applications

Trauth emphasised that “urbanOS was developed in collaboration with several cities and municipalities as part of concrete projects, not created in isolation”. For example, in the city of Dormagen, located between Cologne and Duesseldorf, numerous streetlights are equipped with sensor boxes placed about 3 m above the ground. The sensors inside measure environmental factors such as humidity, CO₂ levels, particulate matter and noise, while also counting pedestrians, cyclists and vehicles passing by. The city’s sensor network helps identify areas with high traffic congestion, provides retailers and restaurateurs with insights into customer flows, and assists in more targeted planning and evaluation of city events. It also identifies areas with excessive heat exposure, allowing the city to protect the population during summer with shading measures to combat the effects of global warming.

The city of Huerth, located near Cologne, has also optimised waste disposal using urbanOS. Public waste bins have been equipped with ultrasonic sensors that measure the fill level and transmit the data wirelessly to the smart city operating system. There, AI determines the most efficient route for the garbage trucks to empty the bins. By dynamically adjusting the previously fixed route, unnecessary trips are avoided, saving about 20% in costs and reducing CO₂ emissions by approximately 30%, while ensuring that the waste bins do not overflow.

According to dataMatters, urbanOS is currently in an early launch phase in over 20 cities, not only in Germany but also in Belgium, France and the Netherlands. “Most municipalities understandably prefer to go public only once the projects are significantly more advanced,” Trauth said.

Smart cities of the future

The global market for smart cities is currently estimated at over $700 billion and is expected to grow to around $4 trillion by 2030. The German smart cities market is currently valued at approximately 8 billion euros and is projected to reach up to 47 billion euros by 2030. “The potential for smart cities is immense, and urbanOS is entering the market at exactly the right time,” Trauth said.

The company’s goal is not only to make individual cities smarter but also to connect smart cities with one another, with urbanOS already being used in pilot projects in the state of North Rhine-Westphalia as part of the funding initiative ‘DatenMarktplatz.NRW’, which aims to build a regional data ecosystem.

“Several cities are establishing their data spaces in an urbanOS hub, where, with the municipalities’ consent, the data can be analysed across cities by AI. This allows experiences to be pooled, and cities can benefit from each other. By using the urbanOS hub, municipalities can save significant costs, as they don’t need to build their own AI infrastructures,” Trauth said.

urbanOS is also prepared for the next generation of actuators. While sensors collect data, actuators are devices that trigger actions, such as controlling traffic lights, managing digital displays, operating irrigation systems or overseeing parking guidance systems. Next-generation actuators could include AI robots that are put to use in municipalities as gardeners, cleaning staff, or assistants for public services and security tasks.

“What may sound like science fiction today will become everyday reality in a few years,” Trauth said.

Image credit: iStock.com/Dragos Condrea