Moving air has become a nuisance for the trucking industry in recent years.
A recent industry report estimated that at any given moment, roughly 35% of all trucks on US freeways were empty of goods.
For example, a truck driver might secure a load to haul from Long Beach, California, to Chicago, but once they drop off the load, they’ll head home with a trailer full of, well, nothing but air.
This problem isn’t just about inefficiency but also cost. Wasted time and fuel mean extra expenses for shippers, which eventually leads to higher prices for consumers. The issue is also related to sustainability: Additional carbon emissions and inevitable road congestion undoubtedly affect our environment.
Uber Freight — a business unit of Uber Technologies — has set out to solve this problem and do it almost exclusively with artificial intelligence. It works a lot like the Uber app does on a smartphone.
With the Uber app, riders are the users and request transport from all available drivers. With Uber Freight, truckers and trucking companies are the users and they can use it to line up different loads so their trucks aren’t running empty for more than a few hundred miles a day.
This way, instead of going from Long Beach to Chicago and back, a truck might bring new loads from Chicago to New Orleans, New Orleans to Houston, and Houston to Phoenix before heading home.
The technology behind this platform uses AI to optimize shipping routes, Uber Freight CEO Lior Ron told Business Insider. He said that this technology could cut a truck’s empty rate to as low as 10%.
“The ultimate goal is to make every mile of a trip a paid mile and make it worth everybody’s while that these guys are out there making deliveries,” Ron said. “We can’t achieve that yet, but we sure can come a lot closer.”
Since the trucking-specific Uber Freight platform launched in 2023, it has used machine learning to pioneer an algorithm that ensures carriers receive up-front guaranteed pricing for trucking and freight.
This algorithm has been used by thousands of companies, including 200 of the Fortune 500s, Ron said. He added that the system had moved more than $20 billion in freight.
“By looking at hundreds of parameters, we’ve been able to make the model accurate enough that it has removed all the friction and back-and-forths of trying to estimate the price of trucking,” Ron said. Those parameters include weather and traffic conditions and road closures.
Uber Freight is also using machine learning to address vehicle routing, a complex issue that involves determining the most efficient route for a vehicle to deliver goods to a set of locations. Here, the issue is not so much avoiding traffic as routing trucks so that their trailers are full more often than not.
By algorithmically designing the optimal route for the truck driver, the company has been able to reduce empty miles by between 10% and 15%, Ron said. This benefits vendors, trucking companies, drivers, and consumers since lower transport costs typically translate to lower product costs.
Freddie Jimenez, the owner of F&J Logistics Inc. in Kansas City, Missouri, said that Uber Freight makes it easier for him to plan his day, find loads that fit his schedule, and keep his wheels moving.
“As a driver, the most important thing is staying on the move. I am not wasting hours waiting or worrying about where the next load is coming from,” Jimenez told BI.
Uber Freight’s technology is part of a broader push among logistics companies to use AI to gain a competitive advantage, said Jose Reyes, a senior director and analyst for Gartner’s supply chain division.
“AI systems can analyze weather, traffic, and road conditions to suggest optimal routes in real-time,” Reyes told BI in an email. “This is a tremendous benefit in not only efficiency but with driver safety, load planning, and dispatching. This application of AI can significantly reduce manual work.”
Chris Caplice, the executive director of the MIT Center for Transportation and Logistics, hosted a webinar on AI in logistics with Ron late last year. There, Caplice said Uber Freight’s technology is an example of innovations in the trucking industry.
“By being trained continuously, the models will learn better routing policies automatically; if a policy shifts, for example, the model will pick up on it, eliminating the need for specialty algorithms,” Caplice said during the event. He added: “AI models generalize well to solve previously unseen problems such as vehicle capacities.”
Uber Freight is also deploying agentic AI to improve efficiency.
This flavor of AI hinges on the ability to use human language to mimic human interactions.
Uber Freight deploys the technology in a customer support center and uses it as the first line of defense against complaints.
Ron said that by dispatching canned messages to certain inquiries, the company expected to reduce its waiting time to 30 seconds from five minutes.
These shortened wait times can help with efficiency by minimizing the time drivers spend dealing with customer service for simple tasks, like receiving a link on their smartphone or a document.
