Support Dashboard Refresh

After two years of running the same fleet monitoring interface, the support team at a regional distribution hub in Perak asked us to rebuild their dashboard. The old screen showed raw AGV telemetry, but operators had to cross-check three separate tabs to answer a simple question: which robot is stuck and why. The refresh was not about adding more charts. It was about cutting the time between an alert and a decision.

We started by mapping the actual shift workflow. Support staff do not need battery curves or motor temperatures in the first view. They need a queue of exceptions, sorted by severity, with a one-line reason and a suggested action. We stripped the interface down to three zones: active missions, stalled units, and a maintenance log that only appears when something requires attention.

The second pass dealt with the hardware side. The warehouse runs twelve AGVs on a mixed floor with manual forklift traffic, so the dashboard had to reflect real-world constraints. We added a floor map that shows each robot's last known position, but only when a unit misses its check-in. That single change reduced the number of false alarms operators chased during the afternoon shift.

Implementation took six weeks, including two weeks of shadow testing alongside the old interface. We kept the legacy screen read-only during the trial so operators could compare both views without risk. The new dashboard went live after the third week of parallel running, and the old screen was retired once the team confirmed they had not lost any visibility into the fleet.

The measurable outcome was straightforward: the average time to identify a stalled unit dropped from roughly nine minutes to under three. More importantly, the support team stopped treating the dashboard as a monitoring tool and started using it as a dispatch board. That shift in behavior, not the interface itself, was the real result of the refresh.