One of the most important potential areas for edge computing is the impact on artificial intelligence (AI) at the edge – and a new report from Omdia explores five key areas of growth.
The analyst firm’s paper, titled ‘Connecting the Dots: AI at the Edge’, noted that the Internet of Things (IoT) is one of the main drivers for ‘AI-at-the-edge’, yet a number of different applications across multiple vertical market also apply.
The report noted five factors that would shape AI-at-the-edge:
- New and emerging use cases: looking beyond smartphones, but more towards use cases from enterprise and industrial sectors
- Diverse, cost-effective silicon solutions: AI has helped drive silicon growth for high-end processors, but going forward a diverse portfolio of processors will be needed to help AI proliferate at the edge
- New compute ecosystem from cloud to edge: technologies such as Kubernetes, to make hybrid and distributed cloud infrastructures more manageable and agile, will be key
- Communication service providers: CSPs’ utilisation of such technologies, achieving low latency for new services, reduced costs and improved customer experience, will be a driver
- Broad range of industry verticals with specific requirements: these range from heavy industries, such as oil and gas, to broader applications across industry in terms of video surveillance
The latter will not be a smooth ride, the report noted, as the ethical implications of analytics for facial recognition, as sister publication AI News has variously reported, are still to be ironed out. Yet other industries are seeing a clear uptick. Healthcare, for instance, has seen an acceleration in AI software development as a result of the Covid-19 pandemic. Some sectors, such as manufacturing and utilities, are at a slower pace.
Plenty has been written on how AI is a killer app for edge computing. Writing for the Enterprisers Project in May, Stephanie Overby noted various interesting areas organisations need to explore. There is a lot of potential to come, as real-time learning at the edge will take time and a minimal amount of AI workflow is edge-based today. Not all problems are a fit for edge AI, Overby added, but all IT infrastructure and architectures should be designed with this in mind as use cases develop.
Edge AI also has the potential to supersede cloud AI, as Bill Morelli, VP for enterprise at Omdia, explained. “AI in the cloud is now a reasonably established technology being implemented by a variety of business processes. As a result of this, there are now a number of compelling use cases within AI that require edge technologies,” he said.
“Capability has finally caught up with the need for AI-at-the-edge solutions,” Morelli added. “This marrying of requirement and technology will drive the space for many years to come across multiple sectors to provide better business outcomes and solutions.”
You can read the full report here (email required).
Want to learn more about topics like this from thought leaders in the space? Find out more about the Edge Computing Expo, a brand new, innovative event and conference exploring the edge computing ecosystem.