Motivation

Take a moment to think about your daily routines – or those that have survived the Covid-19 pandemic – and try to come up with things that bother you on a daily basis. Is it the cold water in the shower, taking what feels like hours to get warm enough so you can step inside without getting a cold shock? Or is it the seemingly endless traffic jam on your way to work, each day on the exact same spot and time?

Now imagine your life if none of these daily hassles existed: Instead of stumbling out of bed in the darkness so you can grab a cozy sweater in the fight against morning chills, you wake up to soft, but vivid lighting and a comfortable temperature, both adapted to your personal preferences. No need to grudgingly wait for the shower to warm up - a smart energy demand system already took care of heating up the water right in time for your usual morning shower.

During your coffee you quickly take a look at your AI-based journey planner app to take the optimal route to work while avoiding congestion. And when approaching your destination at the end of an efficient but relaxed journey, a glance at your smart parking system enables you to directly identify free parking spaces nearby, thereby avoiding the annoying and time-intensive search for a parking space.
Sounds pretty appealing, right?

Application Areas

Smart lighting, temperature and energy systems or AI-based journey planners and parking systems are only some examples of how AI can revolutionize infrastructure and public services.  In fact, the variety of services and facilities that can be made more efficient, safe, reliable, or sustainable with the help of AI is enormous. Here’s a more comprehensive overview of how AI can be leveraged to transform infra-structure and public services.

Transportation:
Applied in self-driving vehicles, route optimization, smart parking systems, traffic management, logistics, autonomous marine and aviation systems, or predictive maintenance, AI has the potential to make transportation safer, more reliable and predictable, more efficient, and more sustainable [3].

Energy systems:
In the energy sector, the main opportunities opened up by AI application include demand and supply forecasting/balancing, integration of renewable energy sources, load management, power generation forecasting, fault detection, power flow optimization and many more [4].

Waste and water management:
AI has great potential to optimize waste management, for instance by forecasting waste characteristics, detecting waste bin levels, routing waste collection vehicles and predicting process parameters and output. In water management, AI can be used for leakage detection and prevention, water quality monitoring, predictive maintenance, efficient water supply and forecasting floods [5].

Air quality monitoring:
Combining IoT and machine learning techniques makes it possible to provide real-time and location-based information of indoor and outdoor air quality. This in turn enables citizens to make more informed decisions about where to spend their time [6].

Agriculture:
AI plays a major role in transforming agriculture, for instance by automated crop monitoring, targeted pesticide application, intelligent irrigation systems, soil sensing, disease detection, or crop demand prediction [7].

Healthcare:
The potential of AI to improve health care is huge, including applications in medical imaging, automated triage, holistic diagnosis, personalized treatment, epidemics prevention and mental illness detection [8].

Public Administration:
With the help of AI, public administration can be enhanced for example by service personalization, fraud detection, automated case management, preventive protection, or policy performance forecasting [8].

Examples of Real-Life Use Cases 

Still not fully convinced of the potential of AI to revolutionize infrastructure and public services? Here’s three examples of how AI has already been applied successfully in these domains.

ParkWise - Intelligent Parking 
ParkWise is a crowd-sourcing parking app which provides drivers, connected and autonomous vehicles with real-time info about free and soon-to-be free street parking spots nearby. ParkWise creates a network of interconnected drivers and vehicles, allowing them to automatically exchange information about parking availability. By leveraging AI and Machine Learning, ParkWise provides parking availability, short-term prediction, parking regulations and other services, including notifications about upcoming street cleaning or snow emergencies at their current parking location [9].

Greyparrot – AI-Driven Waste Recognition Service
Greyparrot offers a complete waste composition analysis solution that automates the manual process of sampling and auditing material through intelligent monitoring and analysis. Through AI-powered computer vision software, Greyparrot provides in-depth data insights to stakeholders in the resources and waste sector - giving them vital information they have previously not been able to access [10].

BreezoMeter - Health-Focused Environmental Intelligence Platform:
BreezoMeter provides real-time, dynamic, location-based air quality data to help people cope with air pollution, an invisible killer that takes almost 10 million lives a year, according to the WHO. The BreezoMeter API provides real-time information covering 5.5 billion people in 67 countries, accurate to within 300 meters. To do that, BreezoMeter enriches air monitoring data with information from a wide range of independent sources, including data from local weather and traffic conditions, air pollution dispersion models, and satellites. The procedure is described in the picture below [11].

Challenges for AI in Infrastructure & Public Services

While AI opens up numerous possibilities to make infrastructure and public services safer, more efficient, more reliable, and more sustainable, there will also be challenges along the way [12].

Security:
It is of utmost importance to ensure the security of our next-generation AI-based urban technologies. Traditional IT security measures are not seldom attacked by hackers already and will therefore not suffice, raising the need for more advanced security methods.

Trust in AI based Systems:
Technology will undoubtedly continue to advance at an astounding pace, but the level of impact it will have on our future cities largely depends on the level of trust individuals and organizations place in these systems. This requires AI technologies that are transparent, explainable, and accountable.

Stakeholder engagement:
To ensure performance, engaging stakeholders into the design processes of AI systems will be crucial. However, frameworks on how to engage diverse stakeholders with varying AI-knowledge into innovative urban design processes are limited.

Conclusion

Despite facing considerable challenges, AI opens up highly promising opportunities for innovating infrastructure and public services in a variety of domains, thereby increasing efficiency, safety, reliability, and sustainability of our services and facilities and – nice side effect - alleviating some of the daily hassles each of us faces in their routines. Inspired by the possibilities and not sufficiently intimidated by the challenges? Then seize the opportunity to get unique insights into the emerging field of AI for infrastructure and public services at the upcoming TUM.ai Makeathon – the track “Infrastructure & Public Services” is just the right fit!

Tickets for the Grand Finale can be found HERE: VIEW GrandeFinale