Improved productivity through the kaizen multiplier effect
Team-led improvement ideas boost safety, efficiency and uptime for a precision molding operation.
Company: A global precision manufacturer of specialty pumps for personal care products.
Objective: Conduct a process mapping to find ways to improve cell layout, machine staffing, safety and productivity for two subassembly machines at one of the company’s sites.
Situation: Two subassembly families are produced on two machines. Each subassembly has eight components and two of these components are shared between subassemblies. There are multiple styles of most components. In all, there are four types of set-ups performed on each machine. A box of finished subassemblies had to be removed from the machine, taped, and palletized every 12 minutes. The machines were producing 93 subassemblies per minute, but the daily history-based goal was 37,200 parts, only 35% of capacity. One operator runs each subassembly machine.
Team orientation: The company selected individuals from all three shifts of the operation to participate on the kaizen team. After conducting lean and kaizen training, DPA senior consultant Mike Beauregard led the team in defining process requirements, running trials, creating workflow diagrams, identifying improvement opportunities and creating a new floor layout.
Findings and challenges: One operator had been assigned to each of the subassembly machines. The kaizen team observed the process and found that the operator’s time was taken up filling component feed hoppers, removing the product, testing, paperwork, and resolving short machine stoppages (microstops).
The team observed an operator running one of the subassembly machines. The size and location of the feed hoppers for each machine meant that they had to be filled constantly, tying up the operators and increasing their travel time.
After they conducted a trial with one operator running both machines, they discovered that the lines were down one-sixth of the time due to microstops. The time losses often happened when the operator was on the other machine so the operator had to recognize the other machine was down and then walk over to it – leading to increased downtime. The floor was very slippery with silicone; one team member slipped and nearly fell during the observation.
After mapping the current state workflow for the two machines and observing these issues, the team generated and began to implement dozens of improvement ideas. They included:
Safety: The team ordered absorbent blankets to prevent the silicone from getting to the floor and are taking steps to increase floor traction in case some silicone still escapes.
Improving the flow and eliminating paperwork: Six of the component feed hoppers are being enlarged to hold one box of components each. The other two components will be converted from boxes to gaylords and the components in the gaylords will be loaded into the machine with autodumpers. Gaylords will also be used on finished subassemblies. Instead of changing boxes every 12 minutes, these gaylords will be changed every 2 to 4 hours. Box by box paperwork will be eliminated by going to a bar-code scanning system.
Streamlining testing: Based on data showing no failures in 3 years and redundant final assembly inspection, the twice per shift functional testing of the subassemblies will be changed to the first-piece inspection only.
Better signaling of microstops: The team also identified some causes of microstops and determined action items to eliminate those. Andon lights will be added to indicate when and where microstops happen in the subassembly machines. Auto-restart will be added to the machines so that a machine doesn’t remain down after it clears its own problems.
The team projected that weekly improvements to and around the machines would yield significant increases in operator availability and production time:
Once the improvements above are fully implemented, they will equal 48 machine-hours per week or one day per machine per week. Once the improvements are fully implemented, the two-machine operation will have the capacity to produce an additional 267,840 subassemblies per week.