In industries, machines don’t always stop because something has broken, rather a lot of unplanned downtime comes from something far more basic: the material wasn’t even there!
At a Tier 1 automotive supplier we worked with, that problem had been building quietly for years. The downtime codes were familiar… missing parts? Waiting for raw material? delayed replenishment? minor stops?
No. The plant couldn’t consistently explain why those delays were happening, and truth be told; that gap matters more than it sounds.
Assessing the Slow Down
Was the picker slow? Did the tugger drift off the route? Had the pallet been sitting in staging since the previous shift? Or had an upstream ‘chokotei’ (one of those brief machine stoppages that nobody logs, but everyone feels) knocked the whole delivery rhythm sideways without anyone capturing it as a root cause?
Before we started the project, answering those questions meant pulling supervisor recollections, reading operator comments, and occasionally running a manual time study. Not nearly enough to explain every shift, every route, every queue. After deploying a location-intelligence-based ANDON solution, the plant reduced machine-starvation-related downtime by an average of 17 hours per week.
The lesson wasn’t about resources at all. Here, the plant didn’t need more forklifts or headcounts first, and instead it needed better evidence.
The Part of the Factory Nobody Was Really Watching
Modern plants are good at measuring machines: cycle time, alarm codes, OEE, scrap— all tracked, logged, analyzed. But the intralogistics layer feeding those machines is a whole different story.
ERP and WMS systems capture transactions: material issued, moved, staged, consumed. What they don’t capture is the physical journey between those events. A tugger that holds standard work and one that quietly drifted off sequence may both appear as completed runs in the logs. Even the forklift queue that backs up at the same choke point every morning will never surface as a root cause.
It’s just background noise in a system that wasn’t built to catch it and the logistics team is left reconstructing causes after the fact.
From Raw Location Data to Actual Operating Logic
The plant already had RTLS data from forklifts and tuggers. That’s a good start but raw location data answers one question: where is the asset right now?
The questions that actually drive decisions are harder:
- Is this route running on time?
- Is this driver ahead of standard work or behind it?
- How long did the vehicle dwell in the supermarket?
- How long has the material been sitting in staging?
- Where do queues keep repeating?
- Which route patterns put the line at risk?
- Of last week’s starvation events and how many were actually caused by logistics?
To answer those, LocaXion has built the fixed-quantity and fixed-time ANDON layers on top of the existing forklift tracking feed, using the plant’s own material-flow logic as the foundation: route standards, timing windows, supermarket locations, staging lanes, and exception rules specific to how that facility actually operates.
Drivers could see mid-route whether they were on track, ahead, or falling behind. Supervisors had the same view and could connect a late delivery to a specific downtime incident without relying on after-the-fact reconstruction.
The conversation shifted. “The line was waiting for parts” became “the delay connects to recurring queue behavior at this one choke point.” That’s a different level of accountability altogether.
Why the ANDON Framing Mattered
The ANDON principle is old. When something drifts outside standard, you find out fast and respond to it. What’s less common is applying that discipline to intralogistics rather than just the machine on the side of the floor.
Building it as a live ANDON (and not just another tracking dashboard) meant:
- Drivers got route-time feedback while the route was still running. Not in a morning report.
- Supervisors received exception views grounded in real movement, dwell times, and queue behavior.
Each downtime incident could be assigned to a specific cause: missing parts, a late milk run, extended staging dwell, route drift, or chokotei-related disruption from upstream.
That level of detail simply had been impractical before. Manual observation could sample the process occasionally, but this measured every route continuously!
What the Plant Manager Remarked
“LocaXion gave us hard intralogistics baselines that we could run daily management on,” described plant manager Patrice Brooks. “Our forklifts and tuggers were tracked and measured continuously, with route time distributions, dwell by zone, queue behavior, staging, and repeatable congestion signatures at choke points. After go-live, the reporting stayed relevant and credible because it was built on real distributions.”
That last point deserves attention. Many plants build dashboards that supervisors quietly abandon after a few months. The numbers stop matching what people see on the floor, and the tool gets ignored. This one held up because it was built on distributions and percentiles, not averages that get smoothed over or gamed.
The same data also fed plant simulation, so route changes, staffing adjustments, and new process rules could be stress-tested before anyone touched live operations.
A Piece to a Puzzle
Remember, RTLS is just one piece and not the whole answer.
This is something vendors don’t always say plainly: location tracking technology is roughly 30 to 40 percent of the solution. The rest is the digital twin, the process rules, the master data, the exception logic, the ANDON workflow, and integration into ERP, MES, WMS, and simulation tools. Without that stack, RTLS is just a dot moving on a screen.
Vendor lock-in is a real risk for exactly this reason. A plant boxed into a single RTLS OEM’s hardware and software eventually ends up fitting its operations to the vendor’s limits, rather than the other way around. UWB makes sense where sub-meter accuracy is non-negotiable, but Bluetooth works well and costs less plus broad coverage, and yet vision-based RTLS catches what tags miss entirely. Vendor-neutral system integrators exist to keep the architecture serving the operation, not the vendor’s roadmap.
Where to Start
Pull one month of downtime records. Find the hours coded as missing parts, waiting for material, delayed replenishment, or unexplained minor stops. Then ask how many of those incidents have a proven logistics root cause with actual evidence behind it and not just a supervisor’s best guess. Most plants will inevitably find a large gap.
A pilot doesn’t need to cover the whole factory. One high-value line, one replenishment loop, and two metrics: starvation hours recovered per week, and the share of incidents carrying a specific, evidenced reason code. The first number proves value. The second proves that the system is telling the truth!
