Installing a management operating system in a 550-person plant

The situation

The plant ran a day shift and a night shift on production, a separate sanitation shift, and maintenance split across three shifts to cover the clock. Roughly two dozen production supervisors reported through department supervisors to three superintendents, with parallel structures in sanitation and maintenance.

The stated need was structured management operating system implementation in both maintenance and production/sanitation, plus supervisor and leadership development — structure and discipline. In plainer terms: shifts handed off without transferring information, performance data arrived too late to act on, and supervisors had never been taught how to run a meeting that changes an outcome.

The diagnostic

Before proposing anything, I ran a structured area analysis across every department. This is a written instrument, not a walkthrough — a standard sequence of questions asked of every area so the answers are comparable:

• Facility and organisation: work centres, direct-to-indirect labour ratio by supervisor and skill, maintenance personnel assigned by trade and shift, actual supervisor start and stop times versus scheduled, break structure and whether stations are relieved
• Planning, scheduling, reporting: what supervisors plan daily, weekly, and monthly; whether a volume forecast exists by line, shift, and day; whether standards exist, whether they are valid, and when they were last updated and by whom
• Capacity and flow: work centre capacity, current output, reject and rework rate per work centre, documented workflow schematic by product
• Changeover and maintenance interface: current changeover standards and how they are measured, where SMED applies, how maintenance learns about equipment problems, whether work orders carry estimated versus actual times

Alongside this I ran systems mapping with the client's own teams — current management operating system, current maintenance MOS, MRO parts availability and criticality, and sanitation coordination — plus goal and strategy alignment from the top floor to the shop floor, and a supervisor opinion questionnaire to read leadership tendencies.

What was built

Structured shift transition

A daily transition meeting between every outgoing and incoming supervisor, facilitated by the outgoing supervisor, with a fixed agenda: safety incidents and near misses; quality and USDA holds and concerns; then hard current-shift numbers — downgrade, miscuts, bone-in tender, inventory, per-line tender goals against outcome with barriers named, and x-ray infeed and bone-free percentages by machine.

The point of a fixed numeric agenda is that it makes an incomplete handoff visible. A supervisor who cannot fill in the line cannot pretend the shift went fine.

Daily huddles with a scored audit

Huddles were established per department per shift, on-line with employees where the layout allowed. Critically, I built a management huddle assessment sheet so managers and directors could score huddle quality against fixed conditions — all points or none, no partial credit:

• Did the meeting start and end on time? Was everyone present and on time?
• Was the agenda prepared beforehand?
• Did the supervisor review prior-day metrics — production, lost time, schedule, safety, unaccounted time?
• Did the supervisor control the huddle, and acknowledge crew concerns in a way that builds respect and removes excuses?
• Were metric trend graphs presented and the trends actually discussed?
• Were expectations set for output against schedule, and for each individual job?
• Did the supervisor check for understanding by questioning employees?

Weighting the metric review and the trend discussion at twenty points each is deliberate. Those are the two conditions most likely to be skipped and the two that most determine whether the huddle changes anything.

Short-interval control dashboards

The core argument I made to leadership: financial reporting and trending verifies and confirms actions already taken, but arrives too late to drive short-term improvement. It keeps score. It creates a need to explain why or why not. You cannot change history.

Short-interval control data does something different — it gives supervisors and leads real-time understanding of the results of their own efforts, acts as a visual tool to engage employees in understanding their contribution, raises the question of where to focus and who to recognise, and functions as a predictor of the financial result rather than a report on it.

Deployment was staged deliberately: preliminary dashboard rollout, a defined refinement period, then lock-in with a fixed date after which the dashboard was the operating standard.

Supervisor development

Field coaching on standard work routines, 6S, lost time reporting, data collection, data entry, and report utilisation — followed by a confrontation workshop taking supervisor coaching to the next level. Most supervisors are promoted for technical competence and then expected to hold peers accountable with no training in how to do it. That gap is where MOS implementations usually die.

Gap closure

A key event schedule tracked the difference between the documented "as is" and target "to be" management operating system, with continuing work assigned against each gap — plus root cause analysis on recurring package label malfunctions and pre-startup checklists for weigh & price.

Observations from the floor

Alongside the systems work, direct observation produced a low-hanging-fruit list that required no capital:

• Operators were sourcing rolled paper towels from sink areas, leaving station to make several rolls at a time and causing a perpetual hand-towel shortage for handwashing. A simple fixture on the four affected machines resolves the entire behaviour.
• No secondary job assignment existed for idle time.
• Some departments rotated breaks to keep lines running; most took breaks simultaneously. The former needed to become standard.
• Machines could be tuned during breaks — heating plates that build film deposit could be swapped for refurbished plates and reconditioned offline, supported by a kanban of ready spares in the maintenance room, rather than waiting for breakdown.
• An air compressor water release valve needed repair; it consumed roughly an hour of maintenance department time daily and carried a freezing risk when water entered air lines.
• Taper stations had space for industrial-sized tape rolls but were running small rolls, forcing frequent changes and avoidable downtime.

And the structural finding: there was quality control but no quality assurance. Defects were caught; opportunities for continuous improvement were never investigated and long-standing problems were never root caused. Work instructions did not exist at all — a direct opening for the augmented reality instruction work covered in Case 03.