PBC Linear, a linear-motion component manufacturer in Roscoe, Illinois, uses augmented reality headsets and tablets to turn its veteran machinists' knowledge into step-by-step work instructions that new hires follow on the shop floor.
The company bought Taqtile's Manifest software in March 2020, at the start of the pandemic, when an ageing workforce and a regional hiring shortage left it with more new employees than senior staff available to train them. Previously a new worker shadowed an experienced colleague for weeks before running a machine unsupervised.
Building a procedure starts with recording an operator working through a machine's cycle — start-up, tool touch-off, blow-off and quality checks — while a colleague films and asks questions. That recording is authored into a Manifest template: steps, photos, videos and notes are added at a computer, then the author puts on a headset and attaches 3D markers in the physical workspace. A QR code fixed near the machine acts as the template's "home base", registering the instructions to the equipment so each marker appears at the component it refers to. On the thermoformer, for example, a spatial marker at step 4 sits beside the oven handle and defines the start and end points of the sliding action the step describes.
Trainees new to a machine work through the full procedure in a headset, capturing photographic evidence at critical steps so supervisors can confirm the work was done correctly. Workers returning to a machine they have not run for a while scan in on an iPad instead and step through the same template on the tablet.
PBC Linear reports that the approach cut the time to bring a new employee up to speed from about three weeks to three days, and that roughly 70% of operator training now takes place through the headset. Rather than templating every process at once, the company narrowed its focus to specific shop areas — lathes, mills and grinders — for depth and consistency. Interns build many of the AR flows, including those covering work alongside the plant's collaborative robots.
