The UK is stepping up efforts to move artificial intelligence out of data centres and into everyday devices, with the launch of a new national research network focused on low-cost, low-energy systems.
Led by Nottingham Trent University, the TinyML UK Network brings together researchers and industry partners to accelerate the use of decentralised AI, where models run directly on devices rather than relying on cloud infrastructure. The initiative is backed by UK Research and Innovation through the Engineering and Physical Sciences Research Council, with the University of Southampton and Imperial College London as co-leads.
The shift comes as businesses face rising costs tied to cloud-based AI, including energy use, data transfer and infrastructure demands. Decentralised AI offers an alternative by processing data locally on sensors, wearables and embedded systems, reducing reliance on constant connectivity and lowering operational overheads.
For industries handling large volumes of real-time data, this approach can cut latency and improve system reliability. It also addresses growing concerns around data privacy by keeping sensitive information on-device rather than transferring it to external servers.
The network is expected to act as a bridge between academic research and commercial application, helping organisations adopt smaller, task-specific AI models that are cheaper to run and easier to deploy at scale. These models can operate independently or collaborate across multiple devices, forming distributed systems that do not depend on a single central platform.
Early use cases are already emerging. In agriculture, on-device AI is being used to monitor livestock behaviour and flag health issues earlier. In safety and security, devices can detect unusual movement or sound patterns and trigger alerts without storing recordings, reducing both data risk and storage requirements.
The focus now turns to how these systems are trained, updated and managed across networks of devices. For businesses, the direction is clear: AI is becoming less about centralised power and more about efficient, embedded intelligence that can operate closer to where data is generated.
Leave a comment