Nobody Runs a Single-Vendor Robot Fleet Anymore, and Interoperability Is the New Integration Cost

Warehouse automation grew up in silos, and the procurement pattern reflects it: mobile transport robots bought from the vendor with the best fleet software, picking arms from whoever handled the relevant item geometry best, sorting and conveyance from an established supplier with a long service relationship. Each purchase was individually rational. The cumulative result is a facility running three or four robot populations from different vendors, each with its own fleet manager, its own map of the building, and its own idea of what is happening on the floor. Coexistence is achievable through spatial separation — give each population its own zone and its own hours, and they will not collide. Coordination is a different problem: having the mobile robot arrive at the picking station when the arm is ready rather than queuing, having the sorter know that an inbound tote is delayed, having a blocked aisle recognised by every fleet rather than rediscovered independently by each. The throughput gap between a coexisting facility and a coordinated one is substantial, and it is invisible on any individual vendor's performance figures because each system genuinely is meeting its own targets. The integration work involved is unglamorous and expensive. Reconciling the fact that two vendors' maps of the same building use different origins and units. Deciding which system holds authority when two fleets want the same corridor. Building a single view of the facility state when every subsystem publishes a different event vocabulary at a different cadence. Facilities that budgeted automation as a series of equipment purchases and treated integration as a line item consistently underestimated this, because it scales with the number of vendor pairs rather than the number of vendors. Interoperability standards for fleet coordination exist and have been adopted unevenly — enough for basic traffic and status exchange, not enough for the tighter interleaving that produces the largest throughput gains. The pragmatic pattern emerging among larger operators is to run an orchestration layer they own above the vendor fleet managers, treating each vendor's system as a subordinate resource rather than a peer. That is a meaningful strategic choice as much as a technical one: it makes the operator responsible for coordination logic, and it makes any individual vendor replaceable, which is precisely why the operators large enough to staff it are the ones adopting it.