Autonomous Vehicles for Logistics

Industry Application

Autonomous VehiclesLogistics & Supply Chain

Autonomous vehicles are reshaping logistics and supply chain at every tier of the freight network—from long-haul Class 8 trucking on interstate corridors to last-mile sidewalk robots navigating suburban cul-de-sacs. The logistics industry, with its predictable routes, high labor costs, and acute driver shortage, has become the most commercially viable beachhead for AV deployment at scale.

Why Logistics Is the Leading Commercial Frontier for AVs

The freight industry faces a structural crisis that AVs are uniquely positioned to solve. The American Trucking Associations estimated a shortage of over 80,000 drivers in the US by the mid-2020s, with demographics making the gap worse over time—the median age of a commercial truck driver is 46, and the pipeline of new CDL holders is thin. Driver wages, which can exceed $80,000–$100,000 annually for long-haul operators, represent 35–45% of total operating costs for trucking firms. An autonomous system that can drive 22 hours a day without federally-mandated rest breaks—Hours of Service regulations require human drivers to stop after 11 hours of driving within a 14-hour window—fundamentally changes the unit economics of freight.

Beyond labor economics, the operational domain of highway trucking is well-suited to current Level 4 autonomy. Interstate highways are highly structured environments: lane markings are consistent, there are no pedestrians or cyclists, and the range of unexpected scenarios is far narrower than urban driving. This is why every major AV trucking company has chosen to launch on Sun Belt interstate corridors—Texas in particular—before expanding to more complex geographies.

Long-Haul Autonomous Trucking: From Pilot to Commercial Reality

Aurora Innovation reached a watershed moment in April 2024 when it launched commercial driverless operations on the I-45 corridor between Dallas and Houston, carrying freight for customers including FedEx, Werner Enterprises, and Hirschbach Motor Lines. Aurora's system uses a combination of LiDAR, cameras, and radar organized around its proprietary FirstLight LiDAR, which achieves detection ranges exceeding 400 meters—critical for a vehicle that takes significantly longer to stop than a passenger car. By early 2026, Aurora had expanded its operational design domain (ODD) along additional Texas routes and was accumulating the safety data required for broader regulatory approval.

Waymo Via, Alphabet's commercial trucking division, has been running freight in partnership with J.B. Hunt and C.H. Robinson, focusing on the Texas triangle (Dallas-Houston-San Antonio-El Paso). Plus.ai has taken a hybrid approach, offering a supervised autonomy system—a driver-assist product it calls SuperDrive—that has logged millions of miles with partners including Estes Express Lines and Werner Enterprises, building the data flywheel needed for eventual full autonomy. Kodiak Robotics has focused on defense logistics contracts alongside commercial freight, providing a diversified revenue base while scaling its technology.

Middle-Mile and Regional Distribution: Gatik's Constrained-Route Model

While long-haul autonomy captures headlines, Gatik AI has carved out a compelling commercial niche in the middle mile—the segment connecting distribution centers to retail stores and dark stores over fixed, repeatable routes of 5–50 miles. Gatik's insight was that constrained routes in the middle mile are an even more tractable problem than highway trucking: the vehicle makes the same trip repeatedly, with highly predictable traffic patterns, loading and unloading at known facilities. Gatik operates autonomously for Walmart in Arkansas and Louisiana, for Loblaw (Canada's largest grocery chain), and for various CPG customers. Its Class 3–6 trucks run fully driverless on approved routes—no safety driver—making it one of the earliest sustained commercial driverless operations in North America.

Last-Mile Delivery: The Robot at the Doorstep

Last-mile delivery—the final leg from a local depot to the customer's address—represents roughly 53% of total shipping costs, making it the highest-value target for automation despite being technically harder than highway driving. Two categories of vehicle are competing here: low-speed autonomous delivery vehicles (ADVs) that operate on public roads or sidewalks at speeds under 25 mph, and larger autonomous vans or small trucks operating in the 25–45 mph range.

Nuro has deployed its purpose-built, occupant-free R3 delivery vehicle with partners including Kroger and Domino's in Texas and California. Because the Nuro vehicle carries no human occupants, it can be designed without the safety systems (airbags, crumple zones, seat belts) required for passenger vehicles, reducing cost and weight. Serve Robotics, which spun out of Uber's robotics team, operates sidewalk delivery robots in Los Angeles in partnership with Uber Eats, completing thousands of deliveries per week as of 2025. Starship Technologies has expanded its campus delivery network to universities, corporate campuses, and dense suburban neighborhoods across multiple countries.

Port, Yard, and Warehouse: The Controlled-Environment Advantage

Seaports, rail yards, and distribution center yards represent some of the most mature AV deployments in logistics precisely because they are private, controlled environments that sidestep public road regulations. Autonomous yard trucks (also called hostlers or spotters) move trailers between dock doors and staging areas without the variability of public roads. Outrider Systems deploys electric autonomous yard trucks at distribution centers and cross-dock facilities for customers including Georgia-Pacific and ABF Freight. Inside warehouses, autonomous mobile robots (AMRs) from companies like 6 River Systems (acquired by Shopify, then sold to Ocado), Locus Robotics, and Boston Dynamics' Stretch have become standard infrastructure at scale fulfillment operations. While AMRs are technically distinct from road-going AVs, they share the same AI stack—perception, prediction, planning, control—and represent the most commercially mature expression of autonomous vehicle technology in the supply chain.

Applications & Use Cases

Long-Haul Freight (Hub-to-Hub)

Autonomous Class 8 semi-trucks operate on defined interstate corridors—primarily Texas routes as of early 2026—carrying commercial freight for major shippers. Aurora Innovation's commercial launch with FedEx and Werner on I-45 represents the leading edge of this segment. The AV handles the monotonous, dangerous highway leg; human drivers manage pick-up and delivery at terminals in a "transfer hub" model that keeps drivers off the highway entirely.

Middle-Mile Repeatable Routes

Fixed-route autonomy between distribution centers and retail stores, dark stores, or cross-docks. Gatik AI operates fully driverless on approved routes for Walmart and Loblaw, running the same trip multiple times per day on routes where every turn, loading dock, and traffic pattern is known. The constrained ODD enables earlier commercial driverless operation than open-ended highway driving.

Last-Mile Consumer Delivery

Low-speed autonomous delivery vehicles (ADVs) and sidewalk robots completing the final leg to residential and commercial addresses. Nuro's R3 vehicle delivers groceries and restaurant orders in Texas and California; Serve Robotics' sidewalk bots handle Uber Eats deliveries in Los Angeles. Unit economics improve dramatically at scale—a single operator can supervise a fleet of dozens of robots simultaneously.

Autonomous Yard Management

Electric autonomous yard trucks move trailers within port terminals, rail yards, and distribution center yards—private, controlled environments where public road regulations don't apply. Outrider Systems deploys these at facilities for Georgia-Pacific, ABF Freight, and others. The controlled environment makes this one of the most commercially mature AV applications: no pedestrians, known geography, defined docking positions.

Port Terminal Automation

Automated guided vehicles (AGVs) and autonomous straddle carriers move shipping containers within port terminals. The Port of Rotterdam, Port of Los Angeles, and Port of Qingdao operate autonomous container-handling equipment at scale. Autonomous horizontal transport vehicles reduce cycle time, improve safety by removing humans from container-handling zones, and enable 24/7 operations without shift constraints.

Driver-Assist and Supervised Autonomy

Level 2–3 systems that handle highway driving while a human remains alert and available to intervene. Plus.ai's SuperDrive platform, deployed on Werner and Estes Express fleets, handles throttle, braking, and lane-keeping on interstates, reducing driver fatigue and fuel consumption through more consistent following distances and speed. This segment is generating revenue today and building the data assets needed for full autonomy.

Key Players

Aurora Innovation — Launched the first commercial driverless trucking service in April 2024 on the Dallas–Houston corridor (I-45), carrying freight for FedEx, Werner Enterprises, and Hirschbach. Aurora's FirstLight LiDAR provides 400m+ detection range. Publicly traded (AUR); partnerships with PACCAR (Kenworth/Peterbilt) and Volvo Trucks for OEM integration.
Waymo Via — Alphabet's commercial trucking division, operating autonomous Class 8 trucks in partnership with J.B. Hunt and C.H. Robinson in Texas. Leverages Waymo's robotaxi safety record and the same sensor stack (LiDAR + camera + radar) proven across millions of driverless miles in its passenger service.
Gatik AI — Specializes in the middle mile on fixed, repeatable B2B routes. Operates fully driverless for Walmart in Arkansas and Louisiana and for Loblaw in Canada. The constrained-route model enables earlier commercial driverless deployment than open-ended highway or urban driving.
Plus.ai — Deploys supervised autonomy (SuperDrive) on commercial fleets today, with full autonomy as the roadmap target. Partners include Werner Enterprises, Estes Express Lines, and Amazon. Has logged over 20 million autonomous miles; took a dual-market approach combining US commercial freight with international deployments.
Kodiak Robotics — Autonomous trucking focused on Sun Belt routes and, distinctively, US Department of Defense logistics contracts—providing a government revenue base while scaling commercial technology. Partnered with Knight-Swift and Prime Inc.
Nuro — Purpose-built occupant-free autonomous delivery vehicles (ADVs). The R3 is designed from the ground up for delivery, not passenger transport, enabling a simpler, cheaper vehicle architecture. Partners include Kroger and Domino's; operates in Texas and California markets.
Serve Robotics — Sidewalk delivery robots operating commercially in Los Angeles via Uber Eats partnership, completing thousands of deliveries per week. Spun out of Uber's robotics team; publicly traded (SERV). Expanding to additional cities with a unit-economics model that improves as fleet supervision ratios increase.
Outrider Systems — Autonomous electric yard trucks for distribution centers and freight yards. Deployed commercially at Georgia-Pacific, ABF Freight, and other large shippers. Addresses the highest-injury segment of trucking operations (yard spotting) while generating positive ROI through 24/7 operation and elimination of yard truck driver labor.

Challenges & Considerations

Edge Cases and Long-Tail Scenarios — Highway trucking is tractable in good weather on known routes, but the long tail of unusual scenarios—construction zones with ambiguous signage, debris on the road, emergency vehicle interactions—remains difficult. Each edge case requires additional engineering or policy intervention. The challenge isn't average-case performance (AV trucks now exceed human safety on measured metrics) but guaranteeing graceful handling of scenarios that appear once per million miles.
Regulatory Fragmentation — AV regulation in the US is a patchwork of state laws, with no unified federal framework as of 2026. Texas, Arizona, and California have the most permissive commercial AV frameworks; other states require a human operator, creating geographic constraints that limit the operational design domain and complicate interstate operations. Federal FMCSA rulemaking for driverless commercial vehicles remains in progress.
Liability and Insurance — When an autonomous truck causes a collision, liability assignment is unsettled: the technology vendor, the fleet operator, the shipper, or the OEM may bear responsibility depending on jurisdiction and circumstances. Insurance markets for fully driverless commercial vehicles are nascent; premiums are high and coverage terms are narrow, adding to operating costs and creating a chicken-and-egg problem with scaling.
Weather and Environmental Limits — LiDAR degrades in heavy rain, snow, and fog; camera-based systems face the same limitations humans do plus sensitivity to low sun angles and dirty lenses. Most commercial deployments in 2026 are concentrated in the Sun Belt specifically to avoid winter weather. Expanding to the full US freight network—including the Northeast, Midwest, and Mountain West—requires solving reliability in adverse conditions that are rare in Texas but routine elsewhere.
Labor and Social Displacement — The American Trucking Associations counts approximately 3.5 million truck drivers in the US. While the industry frames AV trucking as addressing a driver shortage rather than replacing existing workers, organized labor (Teamsters, OOIDA) is increasingly active in lobbying for restrictions on driverless commercial vehicles. Managing the political economy of AV trucking—including transition support for displaced workers—is a non-technical challenge that could significantly affect deployment timelines.
Technology-to-Fleet Integration Complexity — Large carriers operate fleets of thousands of trucks from multiple OEMs, with varying vintage, mechanical condition, and onboard electronics. Retrofitting AV systems onto existing assets (Plus.ai's approach) is technically demanding; purpose-built autonomous trucks require capital investment in new iron. The supply chain for AV hardware components—custom LiDAR units, high-performance compute modules, redundant actuation systems—is still maturing, creating procurement and serviceability challenges at scale.