Skip to content

First 90 Days After Robot Cell Go-Live

Robot cell go-live is not the finish line. It is the start of the period where the plant learns whether the cell can survive normal production conditions: imperfect parts, shift changes, maintenance handoffs, operator recovery, SKU variation, upstream shortages, downstream blockage, and the slow loss of project-team attention.

The first 90 days should be managed as an operating review, not a celebration lap. A cell can pass factory acceptance, site acceptance, and the first production run while still being too fragile to copy across the plant.

In the first 90 days after robot cell go-live, track downtime causes, recovery time, manual interventions, operator confidence, gripper wear, nuisance faults, part presentation defects, maintenance response, spare-part use, and whether performance survives multiple shifts. The cell is not scale-ready until support behavior is stable, not merely until cycle time looks good.

Commissioning proves that the robot can run. The first 90 days prove whether the plant can own it.

The difference is important:

Commissioning provesFirst 90 days prove
The robot follows the intended sequenceThe cell survives real product, shift, and staffing variation
Safety devices and handshakes workOperators know how to recover without engineering help
Cycle time is plausibleThroughput is stable after breaks, changeovers, jams, and restarts
The integrator can debug the cellMaintenance can separate robot, tooling, machine, and product causes
Acceptance criteria were metExpansion assumptions are still realistic

A robot cell that needs constant expert attention may still be technically successful, but it is not operationally mature.

Use three reviews instead of one vague post-launch meeting.

ReviewMain questionWhat to decide
Day 30Is the cell running safely and recovering predictably?Fix urgent recovery, training, and nuisance fault issues
Day 60Are downtime patterns and maintenance needs understood?Update spares, PM tasks, operator aids, and root-cause priorities
Day 90Is the cell stable enough to scale or standardize?Approve, delay, narrow, or redesign rollout assumptions

The first review protects production. The second review protects support. The third review protects future capital decisions.

Do not rely only on robot uptime. Robot uptime can look good while the cell loses production through small repeated interventions.

Track:

  • robot faults by category;
  • safety stops and reset causes;
  • gripper, vacuum, fixture, and part-presentation failures;
  • upstream starvation and downstream blockage;
  • manual intervention count and duration;
  • restart time after faults, jams, changeovers, and breaks;
  • operator calls for engineering, maintenance, or integrator support;
  • rejected picks, dropped parts, misloads, bad pallet patterns, or failed inspections;
  • shift-to-shift performance differences;
  • spare-part usage and wear-item replacement.

This data should be simple enough for supervisors and maintenance to use. If the log is too complex, it will decay.

Related page:

The first month should answer: can normal shift personnel recover the cell?

Review:

  • common stop causes;
  • whether HMI messages are understandable;
  • whether recovery steps are documented at the cell;
  • whether operators know when to clear, reset, call maintenance, or escalate;
  • whether safety stops create confusion;
  • whether the robot returns to a known state after interruptions;
  • whether part presentation issues are blamed on the robot when the real cause is upstream.

The most useful Day 30 action is often not mechanical. It is clearer recovery logic and better operator guidance.

Related page:

The second month should answer: is the support model realistic?

Review:

  • wear items used faster than expected;
  • vacuum cup, finger, sensor, cable, hose, and fixture wear;
  • air consumption and contamination issues;
  • calibration drift or vision lighting changes;
  • robot backup, program version, and recipe management;
  • PM tasks added after real production use;
  • whether maintenance has the tools and access needed for common failures;
  • whether the integrator is still being called for issues that should move in-house.

This is also the point where spare-part assumptions should be corrected. A plant should not wait for the first serious outage to discover that a small wear item has a long lead time.

Related page:

The third month should answer: can this cell be a template?

Review:

Scale questionEvidence
Is downtime understood?Root-cause log shows repeatable categories, not vague “robot fault” notes
Can operators recover?Most common stops are handled without engineering support
Can maintenance own first response?PM tasks, backups, spares, and troubleshooting steps are active
Is throughput stable?Output survives shift changes, changeovers, and normal upstream variation
Are parts presented consistently?Gripper and vision issues are not masking upstream process instability
Is integrator dependency acceptable?Escalation is clear and response expectations are realistic
Can the next cell reuse standards?Layout, tooling, HMI, safety, documentation, and acceptance lessons are captured

If these answers are weak, the best decision may be to stabilize the first cell before approving the second.

Watch for these signals:

  • the cell meets cycle time only when an engineer is nearby;
  • operators bypass or avoid the cell during difficult products;
  • faults are logged as generic robot problems;
  • maintenance is unsure whether issues belong to robot, tooling, machine, vision, or product flow;
  • spare parts are borrowed from other equipment;
  • integrator response is informal and depends on one person;
  • the cell runs well on day shift but poorly on nights or weekends;
  • the plant starts planning copy cells before post-launch data is reviewed.

These are not reasons to abandon automation. They are reasons to slow the rollout decision until the support model catches up.

What to standardize after the first 90 days

Section titled “What to standardize after the first 90 days”

If the cell is stable, capture the standard:

  • accepted layout and service clearance;
  • gripper or EOAT bill of materials;
  • fixture and part-presentation lessons;
  • HMI messages and recovery screens;
  • safety reset and restart procedure;
  • robot and PLC backup process;
  • operator training checklist;
  • PM and inspection tasks;
  • spare-part list with local stock decisions;
  • acceptance criteria for the next cell.

This standard is how the first cell becomes an asset for rollout instead of a one-off achievement.