Overview
WMG, University of Warwick’s Intelligent Vehicles research group works with a variety of industrial partners to address challenges presented by the concept of autonomous vehicles. Operating within this department is a multi-million pound funded project ‘Cloud assisted real time methods for autonomy’ (CARMA), working in collaboration with Jaguar Land Rover and EPSRC (Engineering and Physical Sciences Research Council).
CARMA is an innovative and challenging 5 year research project aiming to develop and test automated driving technology, based on a distributed control system and with an approach that is said to be enabled by an ultra-low latency and a cloud-based infrastructure accessed through 5G.
The project concentrates on fully autonomous cars as well as the external environment, and not just operation of the vehicle, which uses radar and video sensing to interpret the external environment, road conditions and other road users as well as integrating cloud computing. This involves advanced video analysis, fusion and cross learning. The project is due to be tested on a section of public road near to the entrance of the University, with future plans to use a stretch of 50 mile road between Birmingham and Coventry set up with cameras designed for autonomous vehicle operation.
Project requirements
We were approached by WMG, University of Warwick to provide an industrial grade
GPU computer for use inside a vehicle to operate as an autonomous vehicle control system, preferably with the highest graphics and processing performance possible whilst integrating multiple large capacity storage drives and 4G communication and also being suitable to run a Linux OS.
As
in-vehicle GPU computers suitable for autonomous driving applications are a very new technology, there are limited
products available particularly with the customers required specification.
Custom solution
The
Nuvo-6108GC GPU computer from
Neousys was presented to the customer as a suitable solution for their needs, reported to be the world’s first ‘Industrial-grade edge AI
GPU computer supporting high-end NVIDIA RTX 2080 graphics card’. The
Nuvo-6108GC GPU computer is able to support the heavy power consumption requirements in order to offer reliable GPU computing performance in industrial environments, and has a patented airflow design and multiple storage drives in a compact form factor.
To meet the customer's requirements,
Neousys provided a bespoke card cage assembly for the specialised graphics card to secure it in place and protect against the shock and vibration that is commonplace with
in-vehicle applications.
250W dual slot graphics cards are normally only used in tower/rack PCs due to space and airflow/temperature restrictions, and as the customer also required 15TB of storage drives, Impulse Embedded's in-house engineers were able to build the
GPU computer in the UK as part of our
Build To Order Service (
BTOS) in order to meet the exact requirements provided by the customer.
The fact that all this was incorporated in to an
in-vehicle box PC made the
Nuvo-6108GC from
Neousys the perfect all-in-one GPU computing solution for this innovative autonomous vehicle project.
To find out more about WMG, University of Warwick's autonomous vehicle project click
here.