Edge computing
This POSTnote describes edge computing, the use of computing resources in close proximity to the place where data are processed within a network, and some of the opportunities and challenges associated with its use. It supplements POSTnote 629.
DOI: https://doi.org/10.58248/PN631
Edge computing describes the use of computing resources (including data storage and processing) located close to the devices which generate the data, rather than relaying this data to a remote cloud computer to perform computations. It is suited to applications that require quick computation and use large volumes of data, which would require high bandwidth for transfer to the cloud. It may also offer some privacy benefits as more data can be processed locally rather than being sent to a cloud. However, the distributed nature of edge computing, and the need for many individual devices to connect and interact, poses a number of unique challenges to its widespread adoption.
Key Points
- Edge computing can enable fast data processing and reduces the need to transfer large amounts of data across the network, it can also enhance privacy and enable computation to continue if the network connection is lost.
- Edge computing is considered to be an emerging technology that has not yet reached maturity.
- For some applications, edge computing offers advantages over cloud computing or vice versa. However, for other applications a flexible approach that uses both cloud and edge is the most appropriate.
- Edge computing is expected to contribute to many emerging technologies including smart manufacturing, autonomous vehicles and the increased uptake of smart internet-connected devices.
- It may be challenging to coordinate data exchange and share computational tasks between many different devices on an edge computing network.
- New security technologies are being developed to secure edge computing networks that contain many devices without using too many computational resources.
- Although edge computing can offer privacy advantages, it may become increasingly difficult to determine who has the responsibility for complying with data protection law and whether they are equipped to comply as uptake by consumers becomes more widespread.
Acknowledgements
POSTnotes are based on literature reviews and interviews with a range of stakeholders and are externally peer reviewed. POST would like to thank interviewees and peer reviewers for kindly giving up their time during the preparation of this briefing, including:
- Dr Bill Mitchell OBE, BCS
- Professor Carsten Maple, Warwick Manufacturing Group*
- Professor Dimitra Simeonidou, University of Bristol
- Glen Robinson, Microsoft*
- James Lovegrove, Red Hat*
- Joanna Hodgson, Red Hat*
- Adrian Keward, Red Hat*
- Jonathan Legh-Smith, BT*
- Bryan Stiekes, Google Cloud*
- Ksenia Duxfield-Karyakina, Google Cloud*
- Martin Percival, Red Hat*
- Professor Lilian Edwards, Newcastle University
- Neil Stansfield, NPL
- Nicky Stewart, UKCloud*
- Paul Duncan, NPL*
- Paul Martynenko, POST Board*
- Professor Paul Watson, Newcastle University*
- Professor Reza Nejabati, University of Bristol*
- Rebecca Lucas, RUSI
- Rhiannon Lawson, Government Digital Service
- Richard Ward, IBM*
- Simon Hansford, UKCloud*
- Sneha Dawda, RUSI*
- Dr Stephen Pattison, ARM*
- Tom March, Government Digital Service*
*denotes people and organisations who acted as external reviewers of the briefing.