Uptime Is the Product

Abstract

Deployed robots are fleet assets, and the robotics industry has not yet built the discipline that keeps a fleet earning: what a buyer actually pays for is uptime, not intelligence. The evidence arrived in little more than a year — Amazon committing more than $100 million to a Greenwood, Indiana plant whose job list reads welders, powder-coaters and controls engineers rather than AI researchers; UPS ordering 400 trailer-unloading robots from a maker that had about a dozen in the field a year earlier; Attabotics collapsing with systems still running in customers' buildings; and a new part of the U.S. mobile-robot safety standard, ANSI/A3 R15.08-3, placing in-service duties squarely on the user. This paper maps six gaps in the robot lifecycle — a subscription priced on a short failure record, site readiness quietly handed back to the customer, health data locked in each vendor's portal, too few hands on the machine, vendors that may not outlive their robots, and no resale market at the end — and shows that fleet maintenance solved most of them a generation ago: preventive maintenance by meter rather than calendar, out-of-service criteria, campaign and warranty recovery, cost per operating hour. Inside the failures it finds eight businesses and roles, from multi-platform field service and cross-vendor fleet analytics to orphan-fleet support and loss data for insurers, and closes with three futures through 2032 in which the scarce skill never changes: keeping a mixed population of machines earning past the warranty. The robot that wins will not be the cleverest one; it will be the one still working on a wet Tuesday in February, three years after the demo. The sixth paper in the robotics deployment series: REA·006 structured the capital, REA·008 vetted the seller — this one asks who keeps the robot running after it arrives, at what cost, and for how long.