Lenovo looks to bring AI intelligence to the edge

Organizations are increasingly deploying edge computing capabilities for a growing number of use cases where AI will play a pivotal role.

Among the many vendors eyeing a piece of the artificial intelligence (AI) edge pie is Lenovo. It recently announced the availability of its ThinkEdge SE70 hardware, which is based on the Nvidia Jetson Xavier NX platform and integrates with the Amazon Web Services (AWS) Panorama service for AI computer vision (CV) processing functionality.

“Edge AI computing applications are numerous and growing rapidly as businesses intelligently transform operations and services with increasingly powerful analytics and automation capabilities,” said Blake Kerrigan, general manager of the Global ThinkEdge Business Group at Lenovo.

The growing market for AI at the edge

Analyst firm IDC has forecast that by 2023, 70% of Internet of Things (IoT) deployments will include AI technologies for autonomous or edge decision-making, supporting organizations’ operational and strategic agendas.

Mordor Intelligence estimates AI in the IoT market will grow at a 27% compound annual growth rate (CAGR) through 2026. The growth will be driven by applications relying on machine learning and natural language processing (NLP) to automatically identify patterns and detect anomalies in data collected from smart sensors and devices. In an IDG survey, 96% of respondents commented that they believe CV has the potential to boost revenue, while 97% said the technology will save their organization time and money.

Among the primary use cases for AI at the edge is computer vision (CV), which is a capability that enables systems to understand and interpret visual data. CV is used for security, safety, manufacturing quality control and improving retail experiences.

Kerrigan noted that CV applications are numerous and varied. In industrial and manufacturing scenarios, CV can be used to analyze CCTV video streams to detect noncompliance with Personal Protective Equipment (PPE) requirements, identify and avert collisions of people, vehicles and machinery, as well as monitor exclusion zones to prevent accidents.

In a food supply chain scenario, CV is being used to monitor the health of livestock to detect potential disease, so affected animals can be separated early from the herd and protect the population.

“There really is so much potential for CV applications at the edge, and with the ROI improving so much, we’re really going to see a lot more in the near future,” he said.

Part of Lenovo’s broader AI at the edge strategy

Kerrigan’s view is that what worked on the edge relatively well just a few years ago simply will not cut it now.

“Building an enterprise edge network on a patchwork of gateways and single-board computers is no longer good enough,” he said. “For reliable performance at the edge, enterprises are looking for purpose-built devices that are rugged, powerful and secure, as well as at a price point ready to scale. “

Kerrigan emphasized that Lenovo’s strategy is focused on what it views as the next evolution of Edge computing.

“For years, IT decision-makers have been focused on consolidating workloads for cost optimization, and now we are seeing new stakeholders in the enterprise who are dreaming up new and exciting use cases that unlock powerful new insights within their business,” he added. “The SE70 is a first of its kind offering from Lenovo that enables this next evolution of vision inference applications at the edge.”

Vision inference, the ability to derive value from items that can be seen, is what computer vision (CV) is all about and is a key feature in the new Lenovo platform. A version of the ThinkEdge SE70 is offered with AWS Panorama Device software development kit (SDK) pre-installed, allowing enterprise users to transform regular IP cameras into ‘smart’ cameras that run CV apps at the edge.

“AWS Panorama enables tasks to be automated that have traditionally required human inspection to improve visibility into potential issues or deficiencies,” Kerrigan said.

For example, he noted that AWS Panorama can be used to track assets for optimizing operations in facilities and monitoring inventory levels on retail shelves — even in environments with limited internet bandwidth.

“The future is about the smart edge,” Kerrigan said. “Businesses relying on critical deployments at the edge are scaling large, sometimes global, operations.”