Autonomous trucks in US supply chains: what Canadian dock ops should watch

The PepsiCo-Gatik deal is not about robots replacing people. It's about running routes no one wants.

PepsiCo and Gatik announced a multiyear autonomous vehicle contract targeting "hard to staff" segments of the food and beverage network. Translation: short-haul milk runs, late-night lanes, warehouse-to-DC transfer routes that pay $18/hour and burn through drivers in 8 months. This is not full autonomous trucking. This is fixed-route, low-velocity drayage on roads both companies have mapped for years. That context matters, because it changes what actually happens when one of these units rolls into a Canadian dock.

From a warehouse ops standpoint, the real signal is staffing pressure. If PepsiCo is contractually committing to autonomous drayage on "hard to staff" routes, it means driver retention in North American 3PL is broken enough that a capital-intensive bet on robotics beats another round of wage hikes. That's not a technology story. That's a labor shortage story wearing a self-driving car disguise.

Why this matters to Canadian 3PLs and the Port of Montreal

Canada's drayage corridor from the Port of Montreal inland to the 401 has been running tight for three years. Driver turnover in short-haul work sits around 40-50% annually across the region, and recruiting owner-operators for sub-400-km lanes is getting harder. A 2023 Transport Canada report flagged driver availability as the single biggest capacity constraint on containerized freight within the 401-Lachine corridor.

If autonomous drayage proves viable on US routes between distribution centers, the technology will move north. Gatik's platform works on predefined corridors with minimal variables: same start point, same end point, same time window every day. The Port of Montreal to a consolid facility in Dorval. The Lachine warehouse to a big-box DC in Mississauga. Exactly the kind of milk-run drayage that autonomous systems can handle today.

The implication is not that drivers disappear. It's that the margin on those specific lanes tightens. If an autonomous unit costs USD 150,000 to deploy annually and operates without driver wage inflation, a traditional drayage company charging CAD 2,400-2,800 per move on a fixed route suddenly has price pressure from a carrier running the same route at CAD 1,800-2,200 per move.

What changes at the dock

When a driverless truck arrives at FENGYE LOGISTICS or any other sufferance warehouse, the paperwork still happens. The CBSA still inspects the seal. The dock still verifies carton count against the PARS release prior to inbound. What vanishes is the human buffer: the driver who waits 20 minutes while you sort a discrepancy, or who calls his dispatcher to extend the drayage window because your dock door opened late.

Autonomous drayage units operate on a timer. They arrive, dock within a 5-minute window, and leave. No flexibility. No "can you hold the load 30 minutes while we finish pick-pack on the previous shipment." This means dock-to-stock SLAs have to be bulletproof. If your average putaway cycle time is 2 hours and the automated truck shows up at 14:30, you need that cargo positioned and released by 14:25. No cushion.

Cross-dock operations feel this hardest. Most cross-dock cutoffs in the 401 corridor sit around 14:00-15:00 for next-day outbound. When drayage is driver-operated, a 15:15 arrival can still make the cut if the driver calls ahead and the dock holds the door. An autonomous unit that arrives at 15:17 either makes the outbound or sits overnight at in-bond storage rates, which run CAD 40-60 per skid per night depending on the facility and the importer's bond tier. That's no longer a dispatch problem. That becomes a cash-flow problem.

The bond and compliance side

Autonomous drayage does not skip CBSA. Each inbound load still triggers pre-arrival review. The broker still files the PARS or RMD. The dock still cross-references the release memo before the truck leaves the gate. Where automation changes the game is in the margin for human judgment.

A driver with a flagged container can call his dispatcher, who calls the broker, who talks to CBSA, and sometimes gets a hold released same-day or pushed to the next morning. An autonomous unit cannot negotiate. If the system says "hold until exam," the hold exists. This puts weight back on the broker side to get clearance documentation perfect on the first pass, because there is no fallback conversation at the dock.

For importers running sufferance warehouse operations through in-bond cargo handling services, the compliance tail now wags the logistics dog. A 2-hour exam hold that used to be absorbed by drayage flex time becomes a hard overnight at the warehouse. That's CAD 40-60 per skid, plus demurrage if the inbound window is tight. Clean CAD filing and zero miscounts go from "nice to have" to mandatory.

When does this hit Canada

The Gatik deal is explicitly US-focused on PepsiCo's continental network. But autonomous drayage platforms do not respect the border. Port of Montreal handled roughly 2.4 million TEU in 2023, with approximately 40-45% moving inland by drayage within the first 72 hours. If the Port's throughput continues to trend upward and drayage driver availability continues to tighten, autonomous drayage from Port gates to 401-corridor warehouses becomes economically viable by 2026-2027.

What matters now is planning dock-to-stock SLAs and cross-dock cutoff windows as if autonomous drayage is already here. If your operation counts on a 20-minute driver buffer at the gate, that assumption is burning away. Build in 10 minutes. Make the release-to-dock cycle 45 minutes or less. If your cross-dock can absorb a 17:00 arrival on a 15:00 cutoff lane, redesign it so 14:45 is the hard stop. The inefficiency feels real today because you are paying for it in human patience. In two years, you will pay for it in storage fees.

The other shoe: autonomous drayage on fixed routes will drive consolidation. Small importers who cannot fill a full FTL to a single DC will find it harder to negotiate custom drayage windows. The economics of 50-pallet milk runs to five different locations will not survive the margin squeeze from autonomous competitors. This pushes more freight into consolidated less-than-truckload (LTL) models, which means more complexity at the dock, higher per-pallet handling costs, and more reliance on 3PLs like FENGYE LOGISTICS to manage the breakbulk and consolidation SLA.

The staffing story underneath

The real reason PepsiCo signed this deal is not that autonomous trucks are cheaper. It is that hiring drivers is impossible. Wages in short-haul drayage have plateaued while operating costs and regulatory burden have climbed. The average owner-operator on the Port of Montreal–Toronto run is over 55, and the pipeline of new drivers is not refilling fast enough. This is not a Canadian-specific problem. It is a North American supply chain infrastructure problem, and it is only getting worse.

For warehouse operators, this creates an odd opportunity. As drayage companies automate the high-margin fixed routes, the unprofitable variable routes (your 8-pallet mixed shipment at 23:00, your emergency same-day inbound at 16:30) will become scarcer and more expensive. The flexibility you took for granted will cost 15-25% more. But the flip side is that your own staffing might become a competitive advantage. If you can run a dock smoothly with reliable putaway times and clean paperwork, drayage carriers will route their human-driven traffic to you, because the math on those lanes works better with a reliable partner.

Related: Vietnam 301 probe: what Canadian importers should expect ...

Related: Industrial real estate boom won't solve your drayage bott...

Related: ABF Freight's 5.9% Rate Hike: What Hits Your Dock in Q2

What to do about it now

Start with dock-to-stock SLA. Measure your actual putaway cycle time by arrival time, day of week, and cargo type. If you are averaging 2.5 hours on a good day and 4+ hours on a heavy day, tighten it to 1.5-2 hours hard. That is the minimum buffer you need when drayage windows get binary (the truck arrives, the dock is ready, or the truck waits at CAD 150-250 per hour demurrage).

Second, audit your cross-dock cutoffs and your ability to hit them. If you have a 15:00 cutoff and you are regularly making it at 14:55, you are 5 minutes from missing an entire day's outbound. Move the cutoff to 14:30 and rebuild the dock workflow to hit it consistently. This is unglamorous and it takes weeks, but it is the only thing that matters when your drayage partner switches to robots.

Third, talk to your broker about CAD filing accuracy and CBSA release patterns. Ask for a 90-day snapshot of holds, exams, and delays on your inbound freight. If hold rates are running above 2-3%, there is a documentation or classification pattern to fix. Every exam hold becomes a warehouse dwell problem the moment drayage loses flexibility.

The autonomous truck is not coming tomorrow. But the staffing crisis that makes it attractive is here now, and it is already baking into drayage economics and dock-window expectations. The warehouse that treats the dock as a cost center and drayage as a friction point will find itself unable to compete when the truck arrives without a driver to absorb delays.