Article sponsored by ADLINK.
Picture a customer at a workplace cafeteria. They place food and a drink on a tray and slide the tray along until they reach the cashier. Only, instead of a cashier, there’s an overhead camera aimed at the tray. In a few seconds, the video footage is analysed and the transaction processed.
This is one of many uses for edge AI and machine vision shown in a new series of videos by Taiwan-headquartered edge computing company ADLINK. The videos provide a snapshot of how edge AI is advancing, and how one vendor is pitching the technology.
ADLINK sees improvements in the accuracy and responsiveness of edge AI-enabled machine vision enabling a wide range of use cases relevant to retail, manufacturing, healthcare, defence and other sectors.
Retail
For example, this video includes an explanation of an “AI shopping cart”, which uses cameras and weight sensors to detect items a customer puts in their trolley. The trolley functions as a checkout – an NVIDIA-based system in the trolley calculates the bill and chooses advertising to display based on each shoppers’ items.
The same video makes a case for using edge AI-enabled machine vision to spot shoplifters, track stock availability on shelves and analyse customers’ movements in shops.
Manufacturing
This ADLINK video shows an AI edge solution inspecting goods as they move past a camera on a production line – image analysis gives a “pass” or “fail” for each item as it passes the camera. The technology is now accurate enough for inspection of contact lenses, according to the company.
ADLINK sees edge AI providing more accurate and faster inspection than “traditional” automated optical inspection systems used by manufacturers.
Workforce monitoring is another application ADLINK is pushing for edge AI. The same video shows an eye-opening demonstration of edge-based machine vision monitoring people as they work on a production line - analysing their posture, whether they’re at their work stations and their work rate.
Health and safety
Safety is another focus for ADLINK. This video shows edge systems checking if people are wearing protective clothing in a safety zone.
In the healthcare sector, ADLINK is pushing edge AI for improving the speed and accuracy of medical imaging.
NVIDIA sees edge AI-enabled machine vision being used to scanning body temperature, detect if people are wearing masks and look out for patients falling over, among other applications. It is also pitching the use of edge AI in contactless displays that allow hospital employees to control software without touching a keyboard or screen.
NVIDIA also sees the technology being used for measuring social distancing and performing contact tracing in hospitals, retail shops and other public areas.
Cities
This video lays out NVIDIA’s case for using edge AI-enabled machine vision throughout cities.
Verizon has used this technology to monitor traffic in real-time. And some stations in San Francisco’s Bay Area Rapid Transit network have used AI-based video analytics to monitor human behaviour in an effort to reduce fare evasion, according to NVIDIA.
NVIDIA’s spokesperson also suggests using this technology in stadiums, including to detect fights in a crowd.
Defence
ADLINK also has its eye on military uses cases for GPU-accelerated AI. This video discusses edge AI helping to provide the “right data at the right time and right place” in the battlefield. It sees faster signal processing as important to collecting actionable intelligence.
The company also pitches the advantages of edge AI systems in autonomous military vehicles.
Underlying technology
These are just some of the use cases covered in ADLINK’s new videos – the company states that “we’re only just scratching the surface of what can be accomplished”.
The videos also feature views on edge AI trends and issues. And they cover underlying edge AI technology. ADLINK sees its approach – which combines its cameras, embedded computing devices and software with NVIDIA’s Metropolis AI platform – making it possible for organisations to deploy edge AI cost-effectively, quickly and reliably.
As organisations push their use of edge computing and the Internet of Things further, they may benefit from keeping these issues in mind.
Watch these videos to learn more about how ADLINK and NVIDIA are tackling these challenges.
