How Penske is using AI for truck maintenance and repairs

• Penske provides transportation services with a fleet of more than 446,000 vehicles.
• The company uses AI platforms that predict problems and recommend effective vehicle repairs.
• This article is part of “Build IT,” a series on digital technology trends that are disrupting industries.

Penske Transportation Solutions is a leader in transportation services, including truck leasing and rental, fleet maintenance, and logistics and supply chain solutions.

The company, which is headquartered in Reading, Pennsylvania, has a fleet of more than 446,000 vehicles. The scale of its enterprise requires highly streamlined and automated workflows within its operations, including vehicle maintenance. To achieve this efficiency, Penske turned to artificial intelligence.

Solving costly problems

Vehicle maintenance issues can be costly for Penske and affect the customer experience, said Sarvant Singh, vice president of data and emerging digital solutions at Penske.

If a truck breaks down miles from its destination, “we have to provide a replacement vehicle,” he said, adding that “the truck has to be towed” and sometimes technicians and parts are not immediately available.

In 2018, to troubleshoot mechanical problems in vehicles, Penske launched a platform called Guided Repair, built with Hitachi, a technology and innovation company. The system uses decades of data that Penske has collected about its fleet and implements “prescriptive AI,” including deep learning models, to make repair recommendations, Singh said.

For example, if a driver notices a check engine light is on and brings the vehicle to a Penske shop, technicians connect it to a probe and download its data. Singh said Guided Repair quickly analyzes the data and generates a recommendation, such as that a particular sensor needs to be replaced.

He said the system, which is now used on most Penske vehicles, has reduced repair time and repeat repairs. It also helped the company collect real-time data from the trucks, including geolocation, performance metrics and fault codes, which identify maintenance issues.

As a next step, Penske wanted to anticipate and respond to maintenance issues before they occur. The company partnered with Hitachi to develop Proactive Diagnostics, a predictive model that uses artificial intelligence and machine learning to flag vehicles with impending mechanical problems so they can be repaired before they fail.

Putting AI to the test

Proactive Diagnostics, which launched in 2021, combines Hitachi’s “robust data science capabilities” with Penske’s “rich data sets,” Singh said, to predict when a truck will fail.

An external device is attached to the vehicles to collect real-time data, Singh said. The technology uses artificial intelligence to analyze fault code patterns and other fleet data and flag problems before they occur, ensuring trucks are repaired quickly.

The device also alerts customers to potential vehicle issues and automatically sends a service request to a designated maintenance facility to help them schedule service, according to the company.

Singh said the system saves drivers time because they can avoid vehicle breakdowns and waiting for replacements, which can be a big operating expense for Penske. By quickly identifying problems and scheduling repairs, the technology also allows customers to stay on schedule and meet delivery deadlines.

Proactive diagnostics is used on about 150,000 vehicles, and Penske plans to expand the technology across its fleet, Singh said.

The results of an artificial intelligence based solution

Last year, Proactive Diagnostics helped prevent about 90,000 trucks from breaking down, Singh said, adding that it also reduced the time it took to repair a vehicle by 15 minutes to an hour.

The maintenance information generated by Guided Repair and Proactive Diagnostics helps technicians work more efficiently, he said. Technicians can accept or reject the information depending on the situation.

“It’s not about replacing humans with machines,” Singh said, adding that AI has been especially helpful for new technicians learning on the job.

Looking ahead

Penske continues to develop AI use cases, Singh said, including building AI tools that allow customers to analyze their own fleets.

In April, Penske Truck Leasing, a Penske company, launched Catalyst AI, an AI platform that allows customer fleet managers to automatically compare their trucks to similar fleets to see how they’re performing based on fuel consumption and other metrics . This is traditionally a manual and time-consuming process that is based on industry benchmarks.

“The transportation industry is being revolutionized by AI,” Singh said, adding that Penske will continue to use emerging technologies to provide better customer experiences and streamline operations.