Brant Nystrom, Lexmark laureate, sat down with Timothy Dooner to discuss the latest developments in Lexmark’s deployment of AI in truck operations.
Lexmark’s new Optra Edge executes AI applications at the point of need to drive greater productivity and savings while delivering a competitive advantage.
According to Nystrom, Optra Edge acts as an end-to-end edge computing platform that brings the potential of AI to business operations. “We enable vision AI to extract information from operating environments and turn that into actionable insights,” he said.
“While we have customers in all kinds of industries, over the past year or two, a large number of folks in trucking, transportation and logistics have taken advantage of our platform.”
Optra Edge provides much-needed solutions to a wide variety of trucking-related issues, including identifying arrival of vehicles, tracking locations of assets within a distribution center or warehouse, yard and gate management solutions, trailer loading and unloading optimization, personal protective equipment enforcement, and a variety of other customizable management systems.
“We attach one of our Edge units to existing camera infrastructure and utilize AI solutions that extract information from those video feeds,” Nystrom said, explaining the product’s implementation.
“We use that information to create data visualizations, provide push notifications and alerts, inform the right individuals of certain events that take place, or send that data to other companies or systems for processing.”
The Optra Edge device management portal allows users to see the status of their devices and tap into real-time views of their operations, enabling solution scalability across hundreds or even thousands of locations if necessary.
“We’re busy in the freight space, but we’re very focused on solving the problems faced daily by this industry,” Nystrom said.
With the dynamic and customizable interface Optra Edge offers, users can get specific solutions for their data-gathering and processing needs. One such solution, Nystrom said, came in the form of a yard and gate management system for a shipper with distribution centers across North America.
“This particular client has been working to replace manned entry gates with digital ones,” Nystrom said. “They’ve had issues, as many others have, with the tracking and managing of arrivals and departures, as well as current locations of trucks and trailers.”
The custom Optra Edge system watches the entry and exit gates with video cameras then extracts visual data from the cabs and trailers as they arrive, providing textual data, particularly truck and trailer numbers.
“We keep track of all of this in a running database,” Nystrom said.
Associated with all of this information is the ability to search and sort by individual data points (including individual tractors and trailers), see their immediate locations, and view the history of their usage. This enables the user to not only know the live location of assets, but also their histories, which can be informative for operational strategies and a variety of other applications.
Additionally, Nystrom says, images are captured at each arrival and departure event so that human employees can visually verify the information recorded in the database.
“As with all of our solutions, it’s coupled with a slick interface that allows users to see the data by almost any metric, whether that’s time of day, vehicle type and many other combinations that allow forensic assessment of the operation,” Nystrom said. “If the camera can see it, the AI can extract it.”
Another client of Optra Edge needed to track trucks as they arrived and departed from particular landfills in order to calculate the average turn time of trucks at each one of these locations.
“The client started with a simple request for a license plate recognition system to match the incoming and outgoing vehicles,” Nystrom said. “The problem with that was that at landfills, a lot of these trucks are covered in mud, obscured by debris or have plates missing. That makes it difficult to have high reliability at any of these sites.”
“As a result, we began to capture all forms of vehicle data,” Nystrom said. “Things like make and model, color, unique markings, carrier information, and so on. We can now reliably identify each individual vehicle and calculate the time they spend at each location more accurately.”
According to Nystrom, that helped the customer save millions of dollars in waiting time and operating initiatives.
“It’s very forward-looking in that this technology can take visual information and extract it into something useful, which will be the foundation for more advanced automated systems in the near future,” Nystrom said. “We’re very excited to be a part of that.”