Bright idea? UK firm pioneers data centres using lampposts

A UK-based company, Conflow Power Group Limited (CPG), is pioneering a novel approach to data centres by integrating solar-powered, AI-enabled computing units into smart lampposts. The firm has signed an agreement with a Nigerian state to deploy 50,000 of these “iLamps,” which combine low-powered AI processors with solar energy to create a distributed data centre network. Each iLamp is equipped with batteries charged by cylindrical solar panels and contains a small NVIDIA chip capable of running AI tasks using just 15 watts of power. CPG chairman Edward Fitzpatrick envisions that when networked together, these iLamps could collectively deliver the processing power of a traditional data centre without relying on grid electricity, offering environmental benefits. The lampposts also function as AI-powered surveillance devices, capable of monitoring parking violations, speeding, and seatbelt compliance. Some units are already operational in a Warwick Hospital car park, providing CCTV and number plate recognition. Plans are underway to expand their use in Nigeria and potentially in Florida schools and local authorities, with features including facial recognition to identify wanted or missing persons. However, experts have raised concerns about the limitations of such distributed systems for demanding AI workloads, emphasizing that these lamppost data centres cannot replace the high-powered facilities needed for intensive processing. Security is another issue, as the units contain valuable components that could be targeted for theft; CPG has designed the chips to become inoperable if removed. Privacy advocates have also expressed apprehension about the surveillance capabilities, particularly regarding potential bias and misuse. CPG states it will deploy the technology only in partnership with relevant authorities and in compliance with laws and regulations. Beyond data processing and surveillance, Fitzpatrick suggests the iLamps could enable interactive public engagement, such as voting through simple gestures detected by the cameras. This innovation reflects ongoing efforts to decentralize computing infrastructure while integrating smart city technologies, though its practical impact and acceptance remain to be seen.