HUMANOIDS
GXO Eyes 1,000 Apptronik Apollo Humanoids, Pending a Good Performance Review
November 12, 2026 · Updated November 13, 2026
#Humanoids #Warehouse #Robots-as-a-Service #Deal
Warehouse humanoids just got a bigger number to live up to. On November 12, 2026, contract logistics giant GXO announced a letter of intent with Austin-based Apptronik that could lead to up to 1,000 Apollo humanoid robots working in its facilities over the next four years.
Before anyone orders a thousand tiny name tags, note the word "could." The 1,000 figure depends on the robots hitting performance, safety and operational targets that both companies have agreed on. Trials expand in live operations from 2027, covering goods-to-person and person-to-goods picking, case and tote movement, packing and sortation. The first North American deployment is planned for late 2027. Think of it as a very long probation period with a very large potential payoff.
The two companies are not strangers. They started working together in 2024 with a proof-of-concept program, and Apollo has already completed a US pilot with GXO that focused on picking small consumer goods. The commercial side is expected to follow a Robotics-as-a-Service model, so GXO would rent capability rather than buy a fleet outright.
The fine print
- Announced: November 12, 2026, as a letter of intent, not a purchase order
- The robots: Apptronik Apollo humanoids, up to 1,000 worldwide over four years
- Timeline: Expanded trials start in 2027, with first North American deployment in late 2027
- Tasks: Picking, case and tote movement, packing and sortation alongside existing automation and GXO staff
- Business model: Robotics-as-a-Service, with rollout tied to agreed productivity, safety and operational KPIs
- Scale: GXO runs more than 1,000 facilities and has more than 150,000 employees
What it means for automation buyers
The structure is the useful lesson here. A big operator is paying for humanoids in stages, with each use case measured against agreed KPIs and a service model that avoids a large upfront purchase. If you are weighing humanoids, ask for the same terms: a narrow pilot, clear pass/fail numbers and a pay-as-you-go option. For tote moving, palletizing and repeatable picking in a stable layout, cobots and AMRs already have known costs and paybacks. Treat humanoids as an experiment to run next to them until the late-2027 results show whether they can match that.