Corvus Robotics Drones Cut GNC Backorders to 98 a Day

At a GNC nutritional supplement warehouse in Whitestown, Indiana, four small drones fly overhead every day. They count more than 2,000 plastic-wrapped pallets nobody wanted to count by hand.

The aircraft come from Corvus Robotics, a startup building autonomous indoor flyers tuned for warehouse inventory work. Since GNC brought the program online two years ago, daily nonshipments have dropped from several hundred to around 98.

The 250,000 sq ft reserve inventory now gets a full audit every month instead of every quarter. One to two people verify the drone data. The original 20-person inventory team has mostly moved to other roles inside the building or left through attrition.

This is what a deployed warehouse drone program looks like in mid-2026. Cheaper than what it replaces, more accurate, and running daily while the trade press still debates whether the use case is real.

How Corvus Drones Run the GNC Reserve Inventory

Bill Monk, GNC’s vice president of distribution, told Business Insider the drones cover specific zones on 25 to 30 minute flights. The schedule is set with Corvus and adjusts around picking activity.

As Business Insider reported, the four drones do not work the entire 250,000 sq ft at once. They cover the reserve inventory, the part of the warehouse where full pallets sit waiting to be moved to active picking.

That is the area where a misplaced box costs the most. A misread location in reserve cascades into a stockout downstream, an order that ships short, and a customer that never sees the product.

Monk’s framing is short and accurate: if you don’t know where it is, you can’t ship it. That is what the drones changed. The inventory system used to think the box was on shelf B-14 when it was actually one slot over. By the time a picker noticed, the order was already missing the product.

Now the discrepancy gets flagged the next time a drone passes that aisle, usually within hours.

AI Onboard the Aircraft and AI Behind the Dashboard

Jackie Wu, the founder and CEO of Corvus Robotics, described two separate AI workloads. One runs onboard the drone for navigation, positioning, and obstacle avoidance.

The other runs above the flight, on the data the drones bring back, and helps GNC staff figure out which discrepancies are worth chasing first. Brendan Englot directs the Stevens Institute for Artificial Intelligence at Stevens Institute of Technology.

He explained why an indoor warehouse is one of the friendliest possible environments for this kind of system. Lighting stays constant. There are few or no windows. The visual scene barely changes day to day. Computer vision models can be tuned to a single building and stay tuned.

That is the opposite of an outdoor pilot fighting glare, weather, and shifting backgrounds.

Englot also noted that drones can offload heavier reasoning to networked cloud models. Less compute on the aircraft means less weight, less battery drain, and longer flight time per charge.

I still remember the first time I flew a drone inside a warehouse. The aircraft was a DJI Inspire 1, and indoors meant no GPS lock, ATTI mode only. I still do not know how I did not crash it. The wall came closer than it should have, more than once. Coordinating flights this precise in narrow aisles is a piece of engineering I respect.

The 600 vs 60 Boxes Moment That Earned the Drones Their Trust

One anecdote Monk shared makes the case for the program better than any spec sheet. A drone reported 600 boxes at a specific reserve location. The GNC inventory system insisted there were 60. Everyone in the warehouse assumed the drone had to be wrong by a factor of ten.

The drone was not wrong. A staff member had typed the wrong box size into the master record and dropped a zero off the end of a measurement.

That is the kind of human input error that quietly poisons a warehouse for months. Without the drone passing that aisle, the discrepancy stays buried until somebody manually counts the slot.

GNC now audits about 2 percent of the drone’s reports for accuracy. Two years in, the audits keep confirming the drone.

Plastic Wrap Is Still the Drones’ Biggest Operational Enemy

The 70 inch (1.78 m) reserve aisles at GNC are narrow. The drones flying them are roughly 46 inches (1.17 m) across, which leaves a slim clearance on either side.

Sensors can see through the pallet’s plastic wrap fine for counting. The failure mode is mechanical. When a worker cuts the wrap to pull a box and leaves a flap of plastic hanging, that flap can catch a spinning rotor. A caught rotor drops a drone.

GNC’s fix is procedural, not technical. Workers are trained to cut and fully remove plastic when they pull boxes. The aisle gets a quick walk-through before any scheduled drone flight. It is the kind of operational tax that does not show up in a vendor slide deck. Real deployments learn it the hard way and bake the fix into the SOP.

The reserve-only scope is the other honest limit. The drones cannot peer into open boxes that pickers have already pulled from, so Corvus offers software-side case-level estimating for those.

Warehouse Workers Got the Boring Job Taken Off Their Plates

GNC’s old inventory team ran 20 people and chronic turnover. The shifts were undesirable. The work was monotonous. Hunting one misplaced pallet in 2,000 is a needle in a haystack with timestamps.

Most of those 20 people moved to customer service, to inventory accuracy in active picking areas, or left through attrition. Nobody got fired into the drones.

Tammy Lacher, a senior specialist in inventory control at GNC, told Business Insider that the drones handle the counting and the team now does more investigative work.

She said it takes a lot of the grind out of the job.

That sentence is the part of the story the AI-replaces-jobs narrative does not have a clean answer for. A boring counting job goes away. A more interesting investigative job stays. Workers who pivot up the value chain end up happier than they were on the old shift.

DroneXL’s Take

Strip away the press release language, and the GNC story is the first warehouse drone case study I have read that sounds like a working program rather than a pilot. Two years in. Four aircraft running daily. Backorders cut from several hundred to ninety-eight. A 2 percent audit rate that keeps confirming the drone.

Workers who actively prefer the new arrangement. The numbers add up because the use case is honest about its limits. Reserve inventory only. Plastic wrap is still a real problem. Narrow aisles still require careful aisle-prep walkthroughs by humans.

What Corvus and GNC found is the same thing every working indoor drone deployment finds. The system has to fit the warehouse’s actual operational rhythm, not the other way around. I expect to see this kind of program running in every serious warehouse in the next ten years.

What I want to see next is how it pairs with the other autonomous floor tools that are coming alongside the drones.

The signal I would watch next is whether one of the warehouse giants, Amazon or Walmart or a 3PL prime, names Corvus or a competitor in a deployment announcement during 2026. That would move warehouse drones from a credible niche into a category that earns a procurement line on every supply chain executive’s roadmap.

Photo credit: Corvus Robotics