Uber Freight Puts a Number on Trucking's Empty-Mile Problem

Uber Freight has spent the past few years drawing attention to a persistent inefficiency in U.S. trucking: empty miles, or miles driven without a load. In a report released in June 2023 and covered by Transport Topics, the company estimated that between 20% and 35% of the roughly 175 billion miles trucks drive annually in the United States are run empty. The report also estimated that drivers could be spending about 3.5 billion hours per year driving without freight.

How Uber Freight arrived at its estimate

According to Transport Topics, Uber Freight's estimate came from comparing total miles driven by for-hire and private fleets against load counts and average haul length, using data from the Bureau of Transportation Statistics and other sources, rather than relying on carrier surveys. The company argued that survey-based estimates, often in the 15%-20% range, understate the problem because they skew toward larger, more efficient fleets. Transport Topics reported that Uber Freight's report also estimated that eliminating empty miles entirely is unlikely, but that reducing them by up to 64% could be possible with broader visibility over loads and trucks.

Uber Freight's approach to routing

Business Insider reported in April 2025 that Uber Freight uses machine learning to help carriers line up sequential loads so trucks are not running empty for long stretches. The article described a hypothetical route in which a truck hauling a load from Long Beach, California, to Chicago could then pick up freight from Chicago to New Orleans, New Orleans to Houston, and Houston to Phoenix before returning. Business Insider reported that Uber Freight CEO Lior Ron said the company's routing technology can reduce empty miles to as low as 10% within its network. Business Insider also reported that the company's pricing algorithm evaluates parameters such as weather, traffic, and road closures to offer carriers guaranteed pricing upfront.

Business Insider also reported that Uber Freight moves more than $20 billion in freight and that its pricing system has been used by thousands of companies, including roughly 200 of the Fortune 500. A separate MIT Sloan Management Review event listing describes Uber Freight as managing more than $18 billion worth of freight.

AI methods discussed at MIT

MIT Sloan Management Review covered a webinar in which Chris Caplice, executive director of the MIT Center for Transportation and Logistics, and Uber Freight CEO Lior Ron discussed different analytic approaches used in logistics. According to MIT Sloan, Caplice distinguished traditional AI (task-focused analytics), generative AI (large language models that summarize and generate content in context), and operations research (optimization methods such as linear programming and network models), and said these approaches can work together rather than replace one another. A separate MIT Sloan article on the same webinar reported that Uber Freight has used machine learning to reduce empty miles from an average of about 30% to between 10% and 15% within its optimized routes.

MarketScale analysis: Companies evaluating routing or logistics AI vendors may find it useful to ask which parts of a proposed system rely on optimization methods with established track records, versus predictive models or generative AI interfaces, since these approaches address different parts of the routing problem.