Venus Pilot

CORE OUTPUT 1

Overall Labour Presence

Of the machines that should be manned right now, the share that are. Machines that are confirmed idle are excluded — nobody is expected to stand at a stopped machine.

0% since shift start

CORE OUTPUT 2

Machine Utilisation — Current Shift

Productive minutes divided by monitored minutes. Resets at each shift boundary.

CORE OUTPUT 3

Shift Changeover Health

Worker presence in the 15 minutes before and after each shift boundary, and any gap between the outgoing and incoming crew.

this shift

Windows analysed --

Windows with gap --

Outgoing (last 15 min) --%

Incoming (first 15 min) --%

Handover gap --

Avg handover presence --%

LIVE

Machine State

What each machine is doing right now, and whether anybody is at it.

LAST 24 HOURS

Top Incidents

The longest alert episodes, ranked by duration. One entry per episode — a machine unsupervised for 43 minutes is one incident, not one per frame.

—

Loading…

LABOUR

Workers by Zone

Live head count per machine. A group is flagged when any machine in it holds more people than its configured limit.

—

Waiting for the first frame…

SINCE SHIFT START —

Monitored i 0m —

Plant utilisation i 0% 0m productive

Labour presence i 0% shift average

Alerts raised i 0 —

BY SHIFT

Shift Comparison

Machine utilisation and labour presence per shift. Bars are drawn from minutes, not from averaged percentages.

—

Machine utilisation by shift

Labour presence by shift

presence coverage

Alerts by shift, by kind

HISTORY

Completed Shifts

—

Shift	Started	Observed	Coverage	Zone D	Zones A+B+C	Presence	Alerts
No completed shifts yet.

Totals are computed from minutes, not by averaging the shift percentages.

Term	Definition	Formula
Machine utilisation	Share of monitored time a machine was actually running. Zone D only -- driven by real carriage-motion tracking. Not tracked for A/B/C, which have no machine-state signal: they report worker presence in their drawn areas only.	productive_minutes ÷ monitored_minutes × 100
Labour presence	Share of monitored time a zone had a worker detected inside its marked ROI region. Real for every zone (A, B, C, D) -- this comes from the worker-detection ROI, never from machine state.	time_worker_present ÷ monitored_minutes × 100
Coverage	Share of the shift's SCHEDULED length that was actually monitored. Low coverage usually means the footage/camera simply didn't span the whole shift, not that the other figures were bad.	monitored_minutes ÷ scheduled_shift_minutes × 100
Overcrowding / Crowding	More workers detected in a zone than its configured limit, sustained continuously for over a minute -- a momentary pass-through doesn't count.	worker_count > max_workers, for > crowding_sustain_sec (default 60s)
Unsupervised	Machine confirmed running with nobody present, for longer than the grace period. Zone D only, since it depends on the carriage-tracked "is it running" signal.	machine running AND worker absent > unsupervised_absent_sec (default 120s)
Needs attention	Count of Unsupervised, Crowding or Changeover-Absent episodes for Zone D in the given period -- the three kinds treated as safety-relevant, restricted to the one zone with full machine-state tracking.	count of episodes, Zone D only
Shift changeover / handover window	The window around a shift boundary (currently configured ±30 min) used to check whether coverage carried through the handover.	boundary_time ± changeover_window_min
Outgoing / Incoming presence	Outgoing = labour presence in the 30 min before the boundary (the crew finishing the shift). Incoming = the 30 min after (the crew starting the next one).	same presence formula, restricted to that half of the window
Gap	Minutes with nobody present anywhere on the line, straddling the handover -- a genuine coverage hole between the outgoing and incoming crew.	minutes where all zones show worker_present = false
Carriage tracking	Zone D's real machine-state signal, from the mill carriage's own motion -- the only machine-state signal the system uses.	velocity ≥ 0.4 px/s = running; ≤ -2.0 px/s = tube-change (backstroke)

Venus Pipes — Unit 1

Zone D — machine utilisation by hour

Zone D — labour presence by hour

Overall Labour Presence

Machine Utilisation — Current Shift

Shift Changeover Health

Machine State

Top Incidents

Workers by Zone

Shift Comparison

Completed Shifts

Venus Pipes — Unit 1

Zone D — machine utilisation by hour

Zone D — labour presence by hour

Overall Labour Presence

Machine Utilisation — Current Shift

Shift Changeover Health

Machine State

Top Incidents

Workers by Zone

Shift Comparison

Completed Shifts

All incidents

Zone legend

Dashboard guide

Load a monitoring CSV

On this machine

Or upload one

Configure Video Source Settings

Video Source & Connection

Local Storage Path