As urban areas continue to grow and evolve, cities are increasingly turning to artificial intelligence to enhance efficiency, sustainability, and overall quality of life.
One of the most promising advancements in this space is edge AI – bringing AI processing closer to where data is generated rather than relying solely on cloud-based solutions. By processing information locally, edge AI offers a range of benefits, including reduced bandwidth usage, enhanced data privacy, improved efficiency, and real-time responsiveness. These factors make it a powerful tool for creating smarter, more responsive urban environments.
Advantages of edge AI for cities
The first question to deal with here is, why bring AI to the edge when there’s already comprehensive cloud-based solutions available?
The first reason is network bandwidth. Sending large amounts of data, like video feeds, across the network for cloud-based processing, consumes a huge amount of bandwidth. While some telecom providers may prefer this because it means more data usage, in many cases – especially in regions where network infrastructure isn’t as strong – it’s much more efficient to process data locally, and edge AI helps reduce that network load significantly.
The second major factor is data privacy. When AI processes data at the edge, it can analyse and even anonymise it in real-time, meaning that potentially sensitive data – such as video footage – doesn’t need to be stored or transmitted elsewhere. Depending on the use case, this is crucial for protecting personal information while still gaining useful insights.
Efficiency is another key advantage. Instead of sending raw data to the cloud for analysis, edge AI allows cities to extract meaningful insights on the spot. In edge AI systems, only relevant metadata is sent to cloud-based systems for further reporting or sharing. This means the system operates more efficiently while still allowing integration with cloud-based analytics when necessary.
Finally, there’s real-time responsiveness. In smart city environments, the ability to trigger actions instantly can be a game changer. For example, in traffic management, edge AI could analyse road conditions and vehicle queues at crossroads and adjust traffic lights immediately rather than waiting for data to be sent to the cloud and processed. In other scenarios, like recycling facilities with fast-moving conveyor belts, real-time AI analysis is essential – there’s no time to send data to the cloud and wait for a response.
| Generative AI at the edge |
|
When we talk about generative AI, the key factor is access to and control over knowledge. Many cities today face budget constraints, especially in light of constantly shifting public funding priorities. Generative AI can help by automating administrative and customer-facing services, reducing the burden on human staff. Imagine an AI-powered assistant that acts as a local knowledge hub – helping citizens navigate permit applications, providing real-time updates on service requests, or assisting field workers with complex tasks by instantly referencing manuals and best practices.
For generative AI, privacy and data security are priorities, which is why cities may prefer to host AI models on local servers rather than relying on a cloud provider. In many cases, this means deploying a localised AI system that can be fine-tuned with city-specific knowledge and run inference directly within municipal data centres. This way, the AI remains within the city’s control, ensuring compliance with data privacy regulations while still delivering the benefits of automation and enhanced user experiences. |
Edge AI in the real world
Edge AI is already proving its value in multiple smart city applications. There are two standout projects that come to mind that demonstrate its effectiveness: smart lighting and smart bus stops.
Starting with a smart lighting project from a PNY partner, AI Tech, one of the biggest outcomes we saw was a 40 per cent reduction in energy consumption. The idea behind edge AI in smart lighting is that AI helps optimise lighting based on real-world conditions. There are two main approaches: one is to maintain a consistent lighting level while still adhering to local regulations, and the other is to dynamically adjust brightness based on factors like weather conditions, road quality, and traffic levels.
This means lights operate at the optimal intensity – bright enough for safety but never using more energy than necessary. By continuously analysing these factors in real time at the edge, the AI can ensure lighting efficiency while cutting down on waste.
Smart lighting also brings smart pole infrastructure to cities, and these can have broader applications where AI plays a role. These can include public and traffic safety insights, people flow and pedestrian analytics to help cities optimise operations, and extend even to smart parking and smart waste management applications – all areas where cities both require more insight into their operations, and where their residents could use more information to make better choices in their daily lives.
For smart bus stops, edge AI is being used to improve both mobility insights and traffic management. One interesting finding in another PNY partner project, led by BusPas, was that about 80 per cent of bus lane violations – like unauthorised vehicles stopping in these lanes – could be detected directly from the bus stop itself. Since bus stops are distributed throughout the city, they serve as valuable observation points, providing continuous real-time traffic insights.
Beyond traffic enforcement, AI at bus stops can also help analyse passenger flow – how many people are waiting, boarding, and exiting buses. For transit operators, this kind of data is really important for being able to optimise transit schedules and improving public transport services. In some cities, like The Hague, passengers check in and out when boarding and exiting, giving transit authorities a clear picture of travel patterns. But in other places, like London, passengers typically only check in when boarding, meaning there’s no precise data on where they exit. AI-enabled bus stops help fill in these gaps by providing anonymised data on passenger movement, making it easier to understand commuting patterns and improve services.
Picking up on the thread of anonymised data in these use cases brings us back to one of the key benefits of edge AI – its ability to enhance data privacy. Because data processing happens locally, video data doesn’t need to be stored or transmitted elsewhere. In some legal contexts, storing viewing video footage from the public sphere requires special authorisation, but simply extracting statistical insights without keeping the footage can often be done without additional regulatory hurdles, which makes edge AI a privacy-friendly solution for urban environments.
Both solutions highlighted here have been designed with urban aesthetics in mind – they don’t require bulky hardware and integrate seamlessly into the urban environment. The computing hardware is built directly into elements like digital bus stop screens, making them visually unobtrusive while still delivering powerful AI-driven insights.
| Other use cases for cities |
There are several promising applications for edge AI in cities, ranging from environmental monitoring to road safety and urban mobility.
|
Understanding and overcoming the challenges of implementing edge AI
Despite its many benefits, deploying edge AI in cities can come with challenges. One of the most significant barriers is a lack of awareness among city officials. Many decision-makers are unfamiliar with how edge AI works and may be hesitant to adopt new technologies due to concerns about data privacy, regulatory constraints, or potential public pushback. Education and outreach are essential to overcoming these concerns. By providing clear information on the capabilities, limitations, and real-world impact of edge AI, cities can make informed decisions about adoption.
Financial investment and infrastructure integration are also critical challenges. Deploying edge AI requires new hardware investments, such as inference devices at the edge. Cities must also determine how AI-generated insights will be integrated into existing infrastructure and governance frameworks. Data storage, access rights, and compliance with privacy regulations must all be carefully considered.
Another hurdle is finding the right technology partners and system integrators. Cities typically rely on public tendering processes, which can be slow affairs. A successful edge AI project requires a well-structured ecosystem – one that includes solution providers, local integrators, and resellers who understand both the technology and the regulatory landscape of a given city or country.
At PNY, we help cities strike the right balance by assessing their infrastructure, needs, and goals. Whether it’s deploying high-performance edge AI solutions for real-time video analytics or implementing localised generative AI models for enhanced public services, we provide the technology and expertise to ensure the best fit for each city’s unique challenges.
We see ourselves as a broker and orchestrator of AI solutions for cities. That means we don’t just offer a single product – we have a broad portfolio of AI solutions and partnerships, allowing us to match the right technology with a city’s specific needs. Our close relationship with NVIDIA is a key advantage here; we’re deeply embedded in their ecosystem, giving us access to cutting-edge AI technologies and a global network of proven solutions.
One of our main strengths is that we don’t just act as a distributor or a hardware provider – we go beyond that. We work closely with cities and system integrators from the very beginning, helping them define their business requirements and understand what has worked in other locations. With our global reach, we can bring in AI solutions that have been successfully implemented elsewhere and adapt them to local needs, ensuring a faster and more effective deployment.
The key to working this way is building strong local partnerships. As a global technology provider, we can bring cutting-edge AI capabilities, but local integrators know how to navigate city regulations, procurement processes, and deployment challenges. That’s why we focus on nurturing an ecosystem – connecting solution providers, OEMs, and integrators through industry events, direct meetings, and relationship-building activities. It takes time, but having the right local partners ensures smoother adoption and long-term success.
In this ecosystem set up, we also take on a consultative role, guiding cities and partners through the complexities of the AI ecosystem. Unlike traditional management consultancies or fully integrated AI solution providers, our expertise comes from real-world deployments, plus an in-depth knowledge of the NVIDIA technology stack. We’ve worked on numerous projects and understand the practical challenges that cities face. That experience allows us to help cities navigate the AI landscape efficiently, connecting them with the right partners, technologies, and best practices.
Embracing edge AI in cities
Edge AI represents a transformative opportunity for cities to become more efficient, sustainable, and responsive to the needs of their residents. By enabling real-time decision-making, reducing operational costs, and improving service delivery, edge AI can drive the next wave of smart city innovation. With the right strategies in place, municipalities can leverage this technology to create urban environments that are safer, cleaner, and more adaptive to the challenges of the future.
---
Autor(en)/Author(s): Youssef Nadiri
Quelle/Source: Smart Cities World, 01.04.2025
Bitte besuchen Sie/Please visit:
