Distributed AI and Machine Learning

Area Overview

Artificial Intelligence (AI) has become a cornerstone of digital innovation, there is a growing need to move beyond centralised models and enable intelligence to be deployed across networks, devices, and locations— this makes AI more scalable, responsive, and privacy-aware.

Distributed AI/ML is essential for enabling real-time decision-making in environments where data is generated and acted upon locally, such as in autonomous systems, smart cities, industrial IoT, and edge computing scenarios.

Traditional centralised AI models often fall short in meeting the demands of modern digital systems, where latency, privacy, and bandwidth constraints are critical. Research into distributed AI/ML is vital to:

Enable real-time, local decision-making without relying on cloud connectivity

Preserve data privacy and ownership by processing data where it is created

Support scalable, collaborative AI systems across organisational and geographical boundaries

Facilitate robust, adaptive learning in dynamic and decentralised environments

Reduce energy and resource consumption through efficient model distribution and updating

This opens up opportunities for more resilient, inclusive, and intelligent systems—capable of adapting to real-world complexity in near real-time.

Our Focus

Nimbus research is focused on addressing the design, deployment and governance of distributed AI/ML systems across infrastructure layers and application verticals. Areas include:

Federated and Collaborative Learning: Privacy-preserving ML model training across decentralised datasets; Creating cross-silo and cross-device federated learning architectures and incorporate incentive models and trust mechanisms for collaborative learning ecosystems. Split learning to enable resource constrained ML model training across heterogeneous infrastructures.

Trust, Security and Explainability: Ensure secure execution environments, tracability of ML workflows and the use of Explainable AI (XAI) in decentralised context to enhance trust in AI decision making.

Trusted AI Workflow Orchestration: Platforms and tools to ensure data provenance and lifecycle management of ML workflows in distributed settings.

At Nimbus, we are building the foundations for intelligent, collaborative, and privacy-respecting AI systems that operate at scale, across sectors and infrastructures

Distributed AI and Machine Learning