Home PC News How BMW and Malong used edge AI and machine learning to streamline...

How BMW and Malong used edge AI and machine learning to streamline warehouse and checkout systems

Watch all of the Transform 2020 classes on-demand proper right here.


During a panel immediately at VentureBeat’s Transform 2020 convention, audio system together with BMW Group’s Jimmy Nassif, Red Hat’s Jered Floyd, and Malong CEO Matt Scott mentioned the challenges and alternatives in AI with respect to edge computing and IoT. While every got here from a special perspective — Nassif from robotics, Floyd from retail — all three had been in settlement that AI has the potential to speed up current work whereas enabling solely new capabilities.

BMW produces a automobile each 56 seconds, Nassif says. Millions of elements circulate into the automaker’s factories from over 4,500 suppliers involving 203,000 distinctive elements numbers, which interprets to about 100 end-customer choices. (99% of orders are fully distinctive.) As BMW’s automobile gross sales doubled over the previous decade to 2.5 million in 2019, this created a logistics dilemma — one which was solved partly by Nvidia’s Isaac, Jetson AGX Xavier, and DGX platforms. Nassif says BMW is tapping them to develop 5 navigation and manipulation robots that transport supplies round warehouses and arrange particular person elements.

“The most important thing we do is bring cars to our customers in the cheapest way we can do it, but with better quality. We need to implement new technology like AI and robotics in order to … support our people on the production line to do their job easier and faster to produce more cars,” Nassif mentioned, including that BMW hopes to have the 5 manufacturing robots in manufacturing by the tip of 2021. “I won’t say it’s easy to convince leadership to adopt these technologies, but it isn’t difficult. They understand the importance of it.”

BMW’s robots — two of which have been already been deployed in 4 factories in Germany — understand the world round them utilizing laptop imaginative and prescient methods together with pose estimation. Courtesy of algorithms skilled on each actual and artificial knowledge, they’re in a position to acknowledge particular elements in addition to folks and potential obstacles (even occluded obstacles) in a variety of difficult lighting situations. To bolster accuracy, the algorithms are frequently retrained in Nvidia’s Omniverse simulator, which BMW engineers from all over the world can log into remotely.

Malong applies machine studying to a special downside: recognizing merchandise at retail self-checkouts. Cameras overhead feed footage of objects on scanning beds to algorithms that spot unintended or intentional mis-scans. Malong’s expertise seems to be for frequent points like occluded barcodes and merchandise left in purchasing carts in addition to “ticket switching,” the place a product is scanned with a less expensive barcode lifted from one other, dissimilar product.

Malong’s algorithms — which run on on-premises Nvidia {hardware} — are skilled utilizing weakly supervised studying, enabling them to be taught to tell apart amongst merchandise from noisy, restricted, and imprecise alerts in video feeds. Once they detect a problem, the corporate’s platform alerts a workers member, who confronts the offending buyer.

Edge computing comes into play right here due to the privateness implications of storing closed-circuit footage within the cloud, Scott says. But edge computing additionally makes Malong’s platform extremely scalable and cost-effective, such that it’s in a position to span 1000’s of shops with out the latency that may be launched by server-side processing.

“Making an AI system scalable is very different from making it run,” Scott mentioned. “That’s sometimes a mirage that happens when people are starting to play with these technologies.”

Regardless of the use case, Floyd emphasised the significance of open platforms with respect to edge computing and AI. Popular frameworks and programming notebooks like TensorFlow and Jupyter are open supply, he famous, as are container orchestration methods like Kubernetes. “With open source, everyone can bring their best technologies forward. Everyone can come with the technologies they want to integrate and be able to immediately plug them into this enormous ecosystem of AI components and rapidly connect them to applications,” he mentioned.

Red Hat facilitates this collaboration with Open Data Hub, a platform-agnostic blueprint full of instruments that deploy an end-to-end AI platform. It’s the muse of the corporate’s personal knowledge science software program growth stack, and it’s designed to assist engineers ideate AI options with out incurring excessive prices or having to grasp fashionable machine studying workflows.

“This allows rapid innovation of new applications and new technologies,” Floyd mentioned.

Most Popular

Square adopts QR codes to bring self-serve ordering to restaurants

Square has introduced a new self-serve ordering feature for restaurants that allows dine-in customers to order and pay for their food through their phones,...

LinkedIn open-sources GDMix, a framework for training AI personalization models

LinkedIn recently open-sourced GDMix, a framework that makes training AI personalization models ostensibly more efficient and less time-consuming. The Microsoft-owned company says it’s an...

Google’s Smart Cleanup taps AI to streamline data entry

In June, Google unveiled Smart Cleanup, a Google Sheets feature that taps AI to learn patterns and autocomplete data while surfacing formatting suggestions. Now,...

ExamSoft’s remote bar exam sparks privacy and facial recognition concerns

Sometimes the light Kiana Caton is forced to use gives her a headache. On top of common concerns that come with taking a state...

Recent Comments