Iowa State University - Home Extension and Outreach - Home ISUE Logo
CIRAS - Iowa State University - Extension and Outreach

CIRAS - Library


Contact us

Continuous improvement of maintenance services at Frigidaire Company

By Jim Black, CIRAS

With increasing capital investment in equipment, most companies are forced to reduce downtime to obtain maximum capacity from the equipment they have. Would your company profit from a 63 percent reduction in downtime cost per unit within six months? This paper addresses how Frigidaire Company obtained results of this magnitude through team processes.

SIPOC Process

When Frigidaire Company embarked on training employees for teams, part of the facilitative leadership training included the SIPOC process developed by Tennessee Associates International in Alcoa, Tenn. SIPOC is an acronym for Suppliers Inputting to a Process which Outputs to Customers.

As manufacturing services manager at the Frigidaire Plant in Jefferson, I decided to form a team to improve maintenance services using the SIPOC process. The team included Doug Wortman - maintenance engineer, John Reineman - 1st shift maintenance facilitator, Rene Holsapple - 2nd shift maintenance facilitator, Chuck Woods - 3rd shift maintenance facilitator, and myself as the SIPOC team facilitator.

After identifying the customers of maintenance, meetings were scheduled for the team to interview these customers to identify their expectations of maintenance and determine what measures they wanted to use to gauge maintenance performance. These expectations were grouped into categories including:

  • Communication with Customers
  • Instill High Expectations
  • Manufacture Machine Parts
  • Establish Mindset/Sense of Priorities
  • Eliminate Repetitive Failures
  • Cross Train Employees
  • Provide Breakdown Communications
  • Assure Sense of Accountability
  • Provide Safety/Environmental Support
  • Develop Customer Measures of Performance and Assure Continuous Improvement
  • Provide Project Support

Actions implemented

Action plans, including what had to be done and by whom, were developed to satisfy each group of expectations. Here are a few examples of the types of actions taken:

Communication with Customers

  • Established shift change meetings for maintenance, production and engineering to review work completed during the last three shifts and to set priorities for the next three shifts.
  • Implemented monthly meetings to review repetitive failures

Develop Customer Measures of Performance

  • Developed charts for measures to track continuous improvement progress on customer measures.
  • Established maintenance meetings to determine best approaches to make most significant improvements in customer measures.

Other outputs of the SIPOC team process included development of a Maintenance Mission Statement, definition of priority codes for maintenance work orders and implementation of a Failure Modes and Effects Analysis process.

Failure Modes and Effects Analysis (sample form available) puts the focus on how equipment might fail and the effects that would be observed to help diagnose the cause of a failure rather than the effect. Corrective action plans are prepared to assure completion of tasks to prevent recurrence of that cause of failure.

Results obtained at Jefferson

Over the next 16-month period, with everyone focusing on customer measures, the following improvements were generated for Work Orders (WOs), downtime and overtime:

Measure Initial level After 16 months % improvement
Open WOs 33 WOs 11 WOs 66%
Average age of WOs 2.00 months 1.19 months 40%
Maximum age of WOs 5.7 months 2.8 months 51%
WO completion 89% 96% 8%
Downtime cost/unit $0.19 $0.07 63%
Overtime cost/unit $0.018 $0.032 -78%0

The cost measures reflect an interesting shift in focus. At the time the SIPOC Team was formed, we had tight overtime controls. When we shifted the paradigm from "minimize overtime cost" to "minimize total maintenance cost," we were able to reduce downtime cost $0.120 while only increasing overtime cost $0.014, i.e. an 857 percent return on the added overtime cost.

Additional efforts at Jefferson included establishing:

  • a training program for mechanics and electricians with three levels of expertise and accompanying pay increases.
  • routine lubrication and inspection tasks for operators to complete.
  • a maintenance management software program which simplified work order processing, improved parts availability, pushed inventory accuracy from 78.0 to 98.5 percent.
  • Kaisen 5-s principles with input from all Maintenance members. The 5-s process includes sort, straighten, scrub, spread and standardize.
  • transition from areactive department, which primarily repaired breakdowns, to a proactive department, which designed and built its own automated handling and processing machinery systems including mechanical, electrical and controls.

Second SIPOC team

Later, as plant engineering manager at the Frigidaire Plant in Webster City, I was asked by Dan Brown, plant manager, to apply a similar approach to improving maintenance services. I formed a team which included Ron Felt - process improvement engineer, Dave Petron - plant engineer, Don Snyder - maintenance engineer, Randy Tempel - 1st shift maintenance facilitator, and myself as the SIPOC team facilitator. After developing our Maintenance Mission Statement, we proceeded with the SIPOC process.

This time, however, we started by flowcharting the process from problem identification to work order generation to assignment and completion of work. Meetings were then held with customers to determine their expectations and measures of maintenance effectiveness. These expectations were numbered and grouped into categories including:

  • Communication with Customers
  • Organization
  • Sense of Urgency
  • Reporting Relationships
  • Excellence of Work
  • Documentation
  • Priority Codes
  • Customer Measures of Effectiveness

Process discussions were held on how each numbered expectation could be addressed. As the Work Order process was improved, the numbered customer expectations were posted to the improved flow chart to show at which stage of the process each expectation could be met. After presenting the team's report to its customers, regular meetings were established to review charts showing performance to the customer measures.

Results obtained at Webster City

After six months, the following results were obtained for Preventive Maintenance (PMs) orders, Work Orders (WOs), and downtime:

Measure Initial level After 6 months % improvement
Open PMs 140 PMs 21 PMs 85%
Average age of PMs 3.8 months 1.0 month 74%
Maximum age of PMs 10 months 3 months 70%
Open WOs 220 WOs 117 WOs 47%
Average age of WOs 1.6 months 1.5 months 6%
Maximum age of WOs 8 months 5 months 37%
Downtime--washer line 0.0175 min/unit 0.0121 min/unit 31%

Additional improvements

Project management methods were applied to shutdown planning. Projects desired by maintenance customers were collected and analyzed for work hour content by maintenance craft. Projects were then posted to the Excel spreadsheet by priority, craftspeople assigned, and the hours for each craftsperson by day scheduled. This made it possible to track the completion of each project (Project Planner Sheet—sample form available), including project specifications and part requirements. The Craftsman Planner Sheet (sample form available) permitted all projects assigned to a specific craftsperson to be posted by day to identify any overlapping (excess) scheduling or open time available.

Another improvement was the implementation of the Oil Leak Identification and Repair History (sample form available) process by Don Snyder, maintenance engineer, in the Plastics Department. The process utilized a two-part tag - one part to be wired to the machine to show the location of the leak and the other part sent to maintenance to initiate a work order. Over a four-month period, 95 percent of leaks identified were repaired. In addition to reduction in the number and magnitude of leaks, the process improved communication between Plastics and Maintenance Departments.

Lessons learned

The first step in implementing change is to raise expectations— acceptance of present levels of downtime or quality is not good enough. Once the opportunities for improvement have been identified, those expected to participate in the continuous improvement process must be trained in the processes to be utilized. Training should include theory and application. Once trained, the teams need time to meet to implement their new processes. Support and follow-up are necessary to answer questions, demonstrate the importance of their efforts, and keep the team focused on tasks that will contribute to improving measures valued by their customers.

Application of Deming's 14 points to maintenance

In the context of Deming's 14 Points for Management, the following summarized how these maintenance efforts fit the Deming principles.

  1. Create constancy of purpose for improvement of products and service. SIPOC is a structured approach that reflects customer needs as continuous improvement to the process.
  2. Adopt the new philosophy. It was no longer possible to accept existing levels of service. The SIPOC process placed focus on satisfying expectations of the customers.
  3. Cease dependence on mass inspection. Not applicable.
  4. End the practice of awarding business on the basis of price tag alone. We had previously contracted some equipment design jobs based on the bid process. See also #8 below.
  5. Improve constantly and forever the system of production and service. See #1 above.
  6. Institute training. The maintenance training program presented materials at three levels:
    • Basic skills—math, reading, hydraulics, and pneumatics
    • Intermediate skills—e.g. gear drives for mechanics and AC controls for electricians
    • Craftsman skills—e.g. machine design for mechanics and robotics for electricians
  7. Adopt and institute leadership. Helping people do a better job and learning by objective methods.
  8. Drive out fear. One of the concerns holding back work order completion was that layoffs would result if the backlog of work got too small. We developed a positive win-win goal: when we get caught up on work orders to the point of extra time available, that time will be committed to designing and building our own process machinery and automation. This was a win for those involved, because they could gain valuable skills by creating first class equipment. It was a win for the company because this equipment would cost less, was developed within our priorities and control, and we retained the expertise upon completion, which improves serviceability of the equipment.
  9. Break down barriers between staff areas. By collaborating on equipment designs, production, engineering, maintenance management, and maintenance craftspeople were able to be simultaneously involved in the process, resulting in improved communication.
  10. Eliminate use of slogans, exhortations, and targets for the work force. (Without methods) by applying a structured process with frequent status reviews, we were able to improve our processes, which, in turn, improved results in our customers' own measures.
  11. Eliminate numerical quotas for the workforce. We avoided setting an arbitrary level of any measure as a quota. Instead we kept asking what it would take to complete a specific group of work orders.
  12. Remove barriers that rob people of pride of workmanship. It was ironic that by protecting the backlog of work (to prevent layoffs), the craftspeople were robbing their customers' perception of maintenance effectiveness. In addition to improving the delivery of completed Work Orders and PMs, the craftspeople were able to take price in the design and manufacture of automating processing and handling equipment.
  13. Encourage education and self-improvement for everyone. In addition to the crafts training referred to in #6 above, salaried members were trained in SIPOC and project management techniques to be able to partner with the craftspeople on major equipment projects.
  14. Take action to accomplish the transformation. By changing the objectives and developing a win-win goal, the team was able to move beyond the paradigm of "an acceptable backlog of work orders" to "let's get the work orders current so we can spend our time designing and building equipment".

--W. Edwards Deming, Out of the Crisis, (Massachusetts Institute of Technology Center for Advanced Engineering Study), 1982, pp 23-96.

CIRAS News, Vol. 30, No. 1, Fall 1995