Many high-value and analytical military tasks, such as object detection, signal classification, anomaly detection, route optimisation, and predictive maintenance, can be performed more precisely and quickly with Artificial Intelligence (AI)/ Machine Learning (ML) models than by any human military analyst, provided the requisite compute power and data pipelines are available where the mission happens.

Whilst Cloud computing has been a major driver of compute power and data storage, it has several limitations in military/defence scenarios. For starters, it relies on high-quality and reliable network infrastructure between the Cloud and the client device(s). This is not a guarantee in case the military’s operational environment lies on the border or in a remote area with little internet coverage—surroundings that represent a large chunk of a military’s combat habitat. On the other hand, storage facilities and computational power are both centralised in a cloud computing environment. Critical data storage and high-level data analysis are routed via a central data centre, causing latency or delays in data processing. This is where ‘Edge Computing’ comes in.

What is Edge Computing?

In technical terms, edge computing is a networking philosophy. Central to this philosophy is bringing computing as close to the source of generated data as possible to reduce latency and network/data pipeline bandwidth usage.[1] Put simply, edge computing decentralises processes of compute, storage and application services from the cloud and transfers the execution of these processes to local places—places closer to where decisions and actions must occur, such as an ‘edge’ server.

The term ‘edge’ is considered fuzzy and refers to a combination of different locations. In the military context, it can refer to either a ruggedised compute on platforms (vehicles, ships, aircraft, unmanned systems), or forward-deployed command posts and mobile ‘mini-datacentres’ or even the processor on a sensor or drone or any other military tool of its kind operating in the battlefield.[2] For example, the BLADe-S is an indigenously developed, AI-integrated wearable language translator that uses edge computing combined with a built-in neural network processor, offering real-time translation of audio, visual and textual intelligence data.[3]

TurbineOne’s Frontline Perception platform is another recent example powered by edge computing and AI. It is used to process complex, multi-source sensor data in environments where........

© IDSA