Production monitoring prototypes

Three small, working concepts that explore what useful shop-floor software could look like for a CNC or auto-component plant. They exist to start conversations, not to sell a product.

Experimental concept — sample data only. Not a live factory deployment. Every figure on this page is invented for demonstration. Nothing you enter is sent anywhere or saved after you leave the page.

Prototype 01

Production monitoring dashboard

A single-screen shift view for a supervisor or production manager: planned vs actual output, downtime and rejections, by hour and by machine. Switch between shifts to see how the same layout highlights different problems.

Shift A · 06:00–14:30

Three-machine CNC turning cell · sample data

Planned production
500
parts this shift
Actual production
420
80 parts below plan
Achievement
84%
Behind plan
Downtime
65 min
all machines
Rejections
12
408 good parts

Hourly output: planned vs actual

The chart needs JavaScript. The same figures are in the table below.

View hourly figures as a table
Hourly planned and actual output
HourPlannedActual
H16058
H26548
H36560
H46056
H56530
H66056
H76558
H86054
Machine-wise summary
Machine Plan Actual Ach. Down (min) Rej. Status
CNC-0118016290%154Watch
CNC-0216011874%383Behind plan
CNC-0316014088%125Watch

Status thresholds used in this concept: On plan ≥ 95%, Watch 85–94%, Behind plan < 85%. In a real plant these would be agreed with the production team.

Research question: does seeing H2 and H5 fall behind during the shift give a supervisor enough time to act? Background reading: why production visibility matters.

Prototype 02

Machine downtime reporting

A stoppage log designed to take seconds: pick the machine, pick a reason from a short fixed list, enter the start time and duration. The summary ranks reasons by lost minutes so the largest loss is always at the top.

Downtime by reason

65 min · 7 stoppages

  • Material not available25 min · 1 stop
  • Tool change18 min · 3 stops
  • Setup / changeover15 min · 1 stop
  • Quality hold7 min · 2 stops

Material not available accounts for 38% of recorded downtime.

Stoppage log · Shift A
StartMachineReasonMinutes
06:20CNC-01Setup / changeover15
07:35CNC-02Tool change6
08:10CNC-03Tool change8
09:05CNC-02Quality hold3
10:15CNC-02Material not available25
11:40CNC-03Quality hold4
12:50CNC-02Tool change4

Entries are kept only in this browser tab and disappear when you leave the page.

Research question: will operators record a reason for every stoppage longer than five minutes if it takes three taps? Background reading: understanding machine downtime tracking.

Prototype 03

Production performance insights

Enter one shift's figures to see achievement against plan alongside availability, performance and quality — the three factors of OEE — plus plain-language observations about where the loss is concentrated.

Shift inputs

Shift length minus planned breaks.

Results

Achievement vs plan
84%
Behind plan
Good parts
408
Availability
85.6%
Performance
98.2%
Quality
97.1%
OEE (A × P × Q)
81.6%

Observations

  • Output is 80 parts below plan (84% achievement).
  • At the ideal cycle time, 65 minutes of downtime equals roughly 72 parts of lost capacity.
  • The weakest factor is availability (stoppages) at 85.6% — the first place to investigate.

Research question: do plant heads find OEE factors useful, or is achievement against plan with downtime reasons enough?

How these figures are calculated

  • Achievement = actual ÷ planned production.
  • Availability = (planned time − downtime) ÷ planned time.
  • Performance = (ideal cycle time × actual parts) ÷ run time.
  • Quality = good parts ÷ actual parts.
  • OEE = availability × performance × quality.

These follow the standard OEE definitions. Achievement and OEE answer different questions: achievement compares output with a plan, while OEE compares it with the machine's theoretical capacity.

Would this work on your shop floor?

Honest feedback from plant heads and supervisors is the most valuable input for these prototypes — including "this would never work here, because…".

Give feedback on a prototype