Machine downtime
Record every stop with its reason, so the same breakdown does not repeat for a year.
The problem
A press stops for forty minutes, twice a day. Nobody writes it down because the operator is busy fixing it. At month end output is short and nobody can say why, so the fix never gets funded.
The flow in Sundial
1. Log the stop where it happens. Create a downtime entry for the workstation with the start and end time and the operator.
2. Give it a reason. Choose the stop reason, such as breakdown, changeover, no material or no operator, and add a note on what happened.
3. Send repairs to maintenance. When the cause is a breakdown, record it against the machine in the asset register so maintenance sees it alongside its schedule and history.
4. Review the pattern. The downtime report totals lost time by workstation and by reason over any period.
What the team sees
Production heads see which workstations lose the most hours and why. Maintenance sees breakdowns tied to the machines it looks after. A request for a new spare or an extra shift comes with numbers, not anecdotes.
Why it holds up
Downtime entries are short to fill in, which is the main reason they get filled in at all. Because every entry carries a workstation and a reason, the totals are comparable across shifts and months. Linking breakdowns to the asset record keeps production and maintenance looking at the same history rather than two separate logs that never agree.
With AI (coming to Sundial)
Sundial is built for questions like “where did we lose the most machine time last month?” in plain language, and for an agent that flags repeating stops and drafts a maintenance task for approval.

