Honeywell International Inc. is a Fortune 100 company with over 100,000 employees and manufacturing sites across the entire globe. They manufacture in four main industries: aerospace, building technologies, performance materials and technologies, and safety and productivity solutions. Honeywell’s Product Lifecycle Management Division is responsible for implementing changes to a products’ manufacturing process in the form of a Component Engineering(CE) project. They operate at sixty four different sites currently. CE projects typically involve switching out a single component of a product for a financial benefit or a functional improvement to the product. To implement a CE project, it must go through the following steps: ideation, engineering, and pre-production. At ideation an improvement idea is created and quoted to check if it can save a minimum of $50,000. At engineering it goes through rigorous sets of testing to ensure functionality is maintained. In pre-production suppliers are vetted and inventory is built up. After this process a project can be successfully implemented.
The current system is operating above capacity at multiple sites. The data showed that 6 of the 64 sites received 56% of the total volume of projects. This uneven distribution has resulted in an increase in the average cycle time. This has also potentially reduced the system throughput.
Based on the initial data that was provided to the team, batching has been detected at various sites. However, according to the client, engineers start on projects at their own discretion and there are no standards for batching at any section in the system.
The project objective is to increase Honeywell’s CE production savings by. By rebalanced utilization across sites overall cycle time can be reduced. By implementing batching guidelines, throughput can be increased.
Through meetings with engineers and processing time data, the team was able to create a simulation model of a sites project flow. Many simulation iterations were generated to create confidence intervals to compare to a site's actual production. When the model was validated, the team created different variations of the model to investigate changes to system performance when batching and parallel processing was implemented.
Using processing times at different stations and the number of workers at each place, a recommended capacity could be calculated. The simulation model was also used to test capacity recommendation.
The Capacity Dashboard allows users to find a site's capacity, cycle time, yearly expected throughput as well as which of the two engineering sites is the system's bottleneck. The user can either upload an excel output from the aforementioned productivity deck, or input the average processing times for each milestone. The user would then have to choose the project type as well as how many engineers of that type exist within the system. With this data the macro will calculate the current system capacity as well as the capacity at maximum utilization before there are projects waiting in queue. While this output may not explicitly make recommendations, it does allow the client to compare the current site to different states. The client can test how changes to the system can affect the work in process, cycle time, throughput, the bottleneck and savings. This will allow the client to make informed decisions about the two engineering sites based on theoretical changes and outputs.
The team tested three changes to the production process at the Engineering step: batching, parallel processing, and both. The team found that the greatest overall reduction in cycle time was shown when implementing Parallel Processing and Batching in tandem. The client can use their own discretion at different sites whether to implement Batching or both Batching and Parallel Processing. Ultimately the lowest project type makeup percentage that allowed for beneficial batching was found to be 25%.
Value and Impact
The dashboard is valuable to the client because it is a tool that can be used to generate current system capacity as well as a recommended batch size to reduce overall cycle time. This dashboard allows the user to input their own data and run different scenarios based on what they put it. Another benefit of this dashboard is that users can see what the effect of increasing the number of workers at a particular site has on the system before they add additional resources. This will tell the user the maximum number of projects a system at a site can handle based on current performance and resources.
The simulation model will allow both the team and the client to validate the results provided by the dashboard. The client will be able to input the site-specific processing times as well as the number of engineering teams for the site-specific project capacity given by the dashboard tool. Running the Simio model for a certain time frame (i.e. one year) will allow the client to run several experiments to gain insight to the average cycle time, throughput, queue times at different stations, as well as confidence intervals for each.
The client is currently experiencing an extremely high volume of projects entering each site with a system average of twenty CE projects per week and in its current state the system cannot support the continuous increase. The capacity tool will give our client a convenient and in-house solution that measures their system capability, to better understand and combat the uneven site project load and recent cycle time and completion rate issues. The joint parallel processing and batching model is a cost-effective and fitted solution for our client that will solidify a consistent batching method across their sites and quicken their testing process. Using our solutions will result in a 12 day decrease in cycle time and 15% increase in project throughput, with the minimum $50,000 cost savings per project this leads to an estimated $4.095 million increase in cost savings.