FedEx and Dexterity just expanded their robotic trailer-loading program at the FedEx Hagerstown Hub in Maryland, moving the system beyond its original pilot site. On the surface, it looks like routine automation news. Underneath, it reveals a strategic choice that separates FedEx from Amazon and could decide which logistics giant scales automation faster over the next five years.
What FedEx and Dexterity Actually Announced
FedEx and Dexterity announced the expanded deployment on July 30, 2026, scaling Dexterity’s Foresight world model and Mech trailer-loading system at Hagerstown. Foresight combines vision, depth, and touch to reason across three spatial dimensions and time, deciding in real time how to place packages for space, stability, and speed. The deployment follows several years of joint development and testing at a smaller pilot site, and it moves the technology into production at a meaningfully larger scale. FedEx says it will also use the Hagerstown rollout to study how the robots affect destination planning, trailer assignments, maintenance schedules, and workforce processes across the hub.
Neither company disclosed how many robots will run at Hagerstown or offered a timeline for expanding to other hubs. FedEx did tell investors at its 2026 Investor Day that it intends to scale trailer loading and unloading across multiple U.S. hubs in the coming years.
The Real Strategy Is Refusing to Build Robots In-House
The more interesting story sits one layer up. FedEx has deliberately chosen not to build a proprietary robotics stack the way Amazon has. Instead, it runs a portfolio of point solutions from outside vendors: Dexterity handles trailer loading, Berkshire Grey’s Scoop bot handles bulk unloading, and the company has previously worked with fulfillment robotics startup Nimble. Stephanie Cook, FedEx’s director of advanced technology and innovation for robotics, told TechCrunch that building this kind of system in-house wasn’t realistic. “We knew it wasn’t something that we could just develop in a matter of months. It was going to take a multi-year journey to get here,” she said of the Berkshire Grey partnership, a description that applies just as well to the Dexterity relationship.
O.P. Skaaksrud, FedEx’s vice president of advanced technology and innovation, framed the logic even more directly in the same interview, arguing that hardware isn’t where FedEx’s real advantage sits. Trucks are just trucks, in his view. The network and software behind FedEx’s delivery system are what make FedEx, FedEx. That’s a variation on a familiar strategic pattern: let outside vendors compete on hardware, and keep the coordinating layer, the actual moat, in-house.
This isn’t FedEx’s first attempt at robotics, and the earlier misses explain the current caution. Its SameDay Bot last-mile delivery robot, launched in 2019, was pulled from New York City sidewalks under then-mayor Bill de Blasio within its first year. FedEx also ended its multiyear partnership with autonomous delivery company Nuro after Nuro shifted its business model away from delivery robots toward licensing its self-driving technology in 2025. Spreading bets across several vendors, rather than committing fully to one, looks less like indecision and more like a company that has already been burned by picking a single partner too early.
Dexterity Isn’t FedEx’s Only Option, and That’s the Point
Dexterity was last valued at $1.65 billion following a $95 million round in March 2025 led by Lightspeed Venture Partners and Sumitomo Corporation, one of its largest customers. Sumitomo has committed to deploying 1,500 Dexterity robots across Japanese warehouses by 2026, giving the startup a second large anchor customer beyond FedEx. But Dexterity isn’t operating in a vacuum. Boston Dynamics’ Stretch, Pickle Robot’s Pickle Unload System, and Anyware Robotics’ Pixmo all compete for the same trailer-loading and unloading contracts. That competitive field matters more than it might seem: it means FedEx can keep leverage over pricing and performance without locking itself into a single robotics supplier, the same optionality that shaped its earlier partnership decisions.
What This Means for the Rest of Logistics
FedEx’s caution about announcing scale, no robot counts, no rollout timeline, tells its own story. This isn’t a company that has solved the economics of warehouse automation and is now racing to deploy. It’s a company still validating unit economics hub by hub before committing capital. That pace sits uneasily next to FedEx’s broader network math: the company closed 93 buildings in 2025 and has flagged at least 24 more closures for the first half of 2026, even as it credits automation and network efficiency with helping it deliver more than 40% of shipments same-day or overnight in the first half of the year. FedEx has not linked the closures directly to robotics, and the two trends may simply share a common driver in cost discipline. But warehouse employees watching robots arrive at a hub that’s also shedding buildings elsewhere have a reasonable basis for asking how the two trends fit together.
Expect FedEx to keep expanding this multi-vendor approach hub by hub rather than betting on one robotics supplier. The strategy trades a proprietary moat for speed and negotiating leverage, and it has already gotten Dexterity’s system to production scale faster than a fully in-house program likely would have. Whether that trade holds up the next time FedEx faces a cost-cutting cycle will say as much about the company’s operating discipline as it does about the robots themselves.
