Bringing AI to the masses: the rising importance of edge infrastructure

AIoTHub (formerly IoT North) Community meetup are pleased to present Omer Rana in "Bringing AI to the masses: the rising importance of edge infrastructure"

Abstract:

Devices proximate to users continue to increase in scale and capacity, including with a number of economic models to own and lease such devices. This edge-infrastructure opens up significant potential for new class of AI applications and algorithms that are "resource-aware". Significant recent work in the distributed systems and the AI community focuses on:

• Integrating edge devices with data centres, especially to enable support for distributed (federated) learning
• Support for caching and/or partial processing of large language models
• Enabling of user devices to be used directly in the inference or learning process.

Scalability in this context needs to consider both cloud resources, data distribution and initial processing on edge resources. This talk investigates how a data analytics and machine learning pipeline can be deployed across the Edge-Cloud continuum.

Understanding what should be executed at a data centre and what can be moved to an edge resource remains an important challenge -- especially with increasing capability of our edge devices and availability of edge accelerators. This talk considers:

1. Partitioning strategies for machine learning algorithms across Edge-Network-Cloud resources based on constraints such as privacy, capacity and device resilience/availability;
2. Mechanisms to “adapt” machine learning algorithms based on the characteristics of devices on which they are hosted. Applications in sustainable transport are used to describe how these strategies can be applied in practice.

Bio:

Professor of Performance Engineering and the International Dean for the Middle East at Cardiff University. He is on the Strategic Advisory Group of UK Research and Innovation (UKRI) International. His research interests include high performance distributed computing (particularly cloud and edge computing) and intelligent systems. Rana has contributed to specification and standardisation activities via the Open Grid Forum and worked as a software developer with London-based Marshall Bio-Technology Limited prior to joining Cardiff University. At MBT he developed specialist software to support biotech instrumentation. He is the “cross-council projects” director for the UK National EdgeAI Hub (https://edgeaihub.co.uk/). He co-leads the “cybersecurity and data science” theme in the UK-US Global Centre for Clean Energy and Equitable Transportation Solutions (CLEETS): https://www.cleets-global-center.org/. Rana holds a PhD in Neural Computing and Parallel Architectures from Imperial College (London Univ., UK). He is the associate Editor-in-Chief of IEEE Transactions on Services Computing and a member of the editorial boards of IEEE Internet Computing magazine and IEEE Transactions on Parallel and Distributed Systems. More details at: https://profiles.cardiff.ac.uk/staff/ranaof