Kanban is a system for moving parts from one lean operation to the next using cards and containers. A translation of the Japanese word Kanban is “signal cards.” Materials and parts are delivered to the point of operation only when a signal is received to release the part to production or to signal upstream operations to build more parts for a downstream operation. The Kanban system is organized into a chain of processes in which orders flow from one process to the next downstream process.
This method allows components to be pulled through the system. In contrast, the old method of production was to push the product through the process to make enough products in case the customer was interested in purchasing something. Equipment failures caused excessive amounts of downtime and the push system produced large amounts of inventory that needed to be stored in warehouses and on production lines, which increased non-value-added costs.
The Kanban system may be simple, but it places undue stress on any plant maintenance and engineering department that isn’t prepared to address scheduled and unscheduled downtime. Minimizing this stress requires one to increase machinery and equipment operational availability, the measure of system readiness to produce when scheduled. Effective Kanban requires addressing scheduled and unscheduled downtime issues either by using design-for-reliability principles in the purchase of more reliable equipment or using design-for-maintainability strategies to make existing machinery and equipment more maintainable. Attacking downtime from either direction enables the production hardware to operate in a Kanban environment more effectively.
| A five-point plan |
The plant maintenance and engineering department that lacks an effective plan for addressing downtime can use these five points to achieve greater maintenance excellence. It’s a process we developed and implemented successfully at many manufacturing locations. The program addresses critical points in deploying a comprehensive maintenance strategy for new and existing hardware in a Kanban environment. It concerns itself with:
Interviewing maintenance personnel and operators can show what measures will reduce the effects of downtime. Unfortunately, you often get the individual’s opinions, not facts about downtime. The interviews are subjective and additional data collection might be needed.
The information obtained from the CMMS is only as good as the raw data residing therein. Most systems only collect data related to maintenance actions, not downtime events. The time shown on a maintenance ticket may not reflect the actual time needed to repair a failure or to conduct scheduled maintenance. The data collected from the CMMS must be filtered to obtain a focused picture of system downtime. Data also can be obtained from your maintenance shops. Having a technician record the time required to make a repair can provide information about necessary, but unscheduled, downtime events, spare parts requirements and scheduled PMs.
The best approach uses a data collector on the floor to monitor equipment operation and gather downtime information. Although costly, it provides the best results when data collection rules are established before Kanban implementation.
The data analysis can be as simple as histograms or Pareto charts listing the downtime events and the labor hours associated with each. Organize downtime into a Top 10 list to guide the activities of the next step.
Low operational availability numbers are a sure sign of impending Kanban system death. Use the data you collected and analyzed to develop a reliability and maintainability plan (R&M) that reduces unscheduled downtime, the number of preventive maintenance interventions and your spare parts inventory. Don’t forget to develop R&M plans for new equipment, as well as existing equipment, to be integrated into the Kanban system.
Before ordering new equipment or machinery, put an aggressive R&M engineering strategy in place to design out potential downtime. Remember, any hardware operating in a Kanban environment must be available when scheduled. It’ll be difficult to obtain desired throughput levels if operational availability is less than 95%.
Machinery designs ought to improve overall reliability. An improvement concept called design for reliability replaces weak design elements with more reliable components. Another way to improve the overall operational availability is to design for maintainability. Addressing these issues early reduces the cost of ownership and makes the Kanban system a cost-effective solution to reducing overall operating costs.
Designing for reliability involves deconstructing a system from its subsystems to its assemblies and down to its lowest replaceable units. Perform a complete reliability analysis on the components to determine the best course of action. You can either replace weaker parts with components boasting a higher reliability immediately, or wait until the component fails. Making this strategic decision allows development of a comprehensive maintenance and spare parts approach later. Addressing design for maintainability issues at this point helps when equipment arrives on the plant floor.
Design for maintainability includes standardizing components, visual markings on production instrumentation and improving maintenance access to components having a high probability of failure. Design for maintainability also encompasses developing a list of standard tools and procedures currently being used by maintenance. This avoids trying to implement new technologies that can be difficult for maintenance technicians to learn.
Because most new Kanban systems are placed in existing facilities, plant maintenance and engineering must attack the same issues applicable to new machinery and equipment. Reliability-centered maintenance requires the same approach as for new designs, but focuses only on the top failure modes. Conducting a machinery failure mode and effects analysis (MFMEA) on selected equipment details failure modes, the effects of failure and the potential causes and risks associated with each failure. The MFMEA also addresses recommendations that reduce the effect of failure on operational availability. Maintenance and engineering can then establish action plans to reduce failures based on a risk assessment and root cause analysis.
Impact of PM, PdM and spares
It’s important to ensure that equipment is available to produce parts on demand. This means plant maintenance and engineering must develop and implement strategies for preventive maintenance (PM), predictive maintenance (PdM) and spare parts inventory using the data obtained above.
PM interventions are considered scheduled downtime events and typically occur when machinery and equipment aren’t producing parts. Needed spare parts kits are preassembled before the PM work begins. This means that maintenance stores must have access to the CMMS if it is to provide a timely response to a PM.
Subjecting hardware to vibration, oil or thermal analysis can detect the potential failures outlined on the MFMEA. PdM intervention ensures that monitoring of bearings, electrical controls, motors and rollers will detect potential failures and avert unscheduled downtime.
The data developed earlier also can assist in optimizing spare parts inventory and identifying a location for storing it. Centralizing spare parts controls critical spare parts and serves as a central warehouse for spare parts distribution. Assembling PM parts kits here ensures that parts are available when needed.
Use satellite spare parts locations throughout the facility for daily spare parts requirements. Feed these mini-cribs from the main stores. Be sure that maintenance personnel account for and report spare parts location and inventory levels to the main stores.
Analyze life-cycle costs
An important aspect in the development of a Kanban system maintenance strategy is life-cycle costs (LCC). The analysis includes costs covering acquisition, operation, scheduled maintenance and unscheduled maintenance.
Reliability and maintainability trade-offs must be completed on issues raised earlier and must include the ability to review whether the design for reliability or design for maintainability strategies should be deployed to address operational availability for the Kanban system. For example, if the data collected suggests that the main cause of downtime is a bearing, the design for reliability strategy could select a large bearing with a greater B10 life. However, the cost for that new bearing could be five times the cost of downtime involved in its replacement. Keeping the original bearing properly lubricated may be the correct solution, and replacing the bearing at failure becomes the replacement strategy. Therefore, the cost trade-off analysis has pointed toward a strategy that also satisfies spare parts and PM issues.
The costs associated with deploying a Kanban system can be kept under control and management then has an objective approach for measuring the effect of maintenance on overall operational availability. Establishing these strategies early ensures successful Kanban implementation.
You can maintain stability in the system you put in place by using a failure-reporting analysis and corrective action system (FRACAS). It’s designed to evaluate production system failures and to implement corrective actions that remove the failure from the process.
Reporting and keeping track of key performance indicators (KPI) applicable to the Kanban system can assist in measuring process stability. It also allows the maintenance and engineering departments to look into the future and determine the best time for PM and to be confident about exactly when failures are likely to occur. When indicators start to change, the maintenance or engineering departments can take corrective actions before downtime affects system throughput. As with any stability investigation, keep your KPIs updated regularly and change them in response to system evolution.
Deploying a Kanban system provides a comprehensive, just-in-time manufacturing process that allows for products to be manufactured with the least amount of waste. It depends on a solid reliability and a maintenance strategy for new and existing equipment to keep waste from overcoming the effects of the Kanban. The five-point process allows for integration and evaluation that ensures your operational availability meets the cost, quality and throughput issues of a lean manufacturing environment.
Douglas R. Malcolm is CEO and technical officer for DRM Technologies Inc. in Sterling Heights, Mich. Contact him at [email protected] and (586) 978-8810.
| 10 Reasons why flow manufacturing implementations have problems|
Failure to dedicate a senior executive to lead the implementation.
Failure to set a business strategy with a clear vision that leverages the capabilities of flow.
Failuer to prioritize the flow initiative in budgeting, target setting and business planning.
Failure to communiticate systematically to emphasize the "why" and how of flow
Failure to provide visible clear leadership that endorses flow.
Lack of supportive manufacturing metrics.
Lack of clear need for urgency
Lack of support for generating short-term wins.
Lack of dedicated resources to support the transition.
Lack of adequate training and execution.