Machine Downtime Tracker.

Build a machine downtime tracker for Crestview Bottling, a beverage co-packer in Columbus, Ohio that runs three packaging lines with nine machines and tracks stoppages on clipboards today. Layout: a persistent filter bar across the top with selectors for line, machine, shift (Day, Night), and reason code, plus a week toggle covering the last six weeks. Below it, four stat cards: Total Downtime (about 96 hours), Availability (93.3 percent), MTTR (48 minutes), and Events Logged (120). The main analytics row holds two charts side by side: a Pareto chart of downtime minutes by reason code, bars sorted descending with a cumulative percent line overlaid, and a line chart of weekly availability with one line per packaging line across the six weeks. Beneath them, a horizontal bar chart ranks all nine machines by total downtime minutes. The bottom region is the event log: a sortable table with date, line, machine, reason code, duration in minutes, shift, operator, and a short note, defaulting to newest first. A prominent Log a Stop button opens a dialog capturing machine, reason code, duration, shift, operator, and note; the dashboard must read perfectly for visitors who never open it. Data: seed 120 downtime events spread across the last six weeks on Line 1, Line 2, and Line 3, with machines named L1 Filler, L1 Capper, L1 Labeler, L2 Filler, L2 Capper, L2 Labeler, L3 Filler, L3 Case Packer, and L3 Palletizer. Reason codes: changeover, jam, starved upstream, blocked downstream, planned maintenance, electrical fault. Durations run 4 to 95 minutes, logged by eight operators such as M. Reyes and D. Okafor split across Day and Night shifts. Plant the story: changeover is the top Pareto reason plant-wide, but the L2 Filler is the outlier machine, generating about a third of all unplanned minutes through repeated jams, and Line 2 availability visibly sags in the weekly trend while Lines 1 and 3 hold steady. Behavior: availability is measured against two 8 hour shifts per line, five days a week. Availability, MTTR, and every chart recompute from the filtered events, so narrowing to Line 2 makes the problem obvious in two clicks. Color reason codes consistently everywhere: planned maintenance neutral gray, changeover in the accent, jams red, starved and blocked amber. All filters combine, and clearing them is one click. Visual identity: use a plum accent for primary actions, chart emphasis, and active states on a clean light neutral base, giving the whole dashboard a focused control-room feel.