Poster
Client Context
Steelcase Inc. is headquartered in Grand Rapids, Michigan and has a diverse portfolio that includes seating, desks, and walls. The Grand Rapids wood plant, our focus within Steelcase’s manufacturing network, focuses on cutting and curating wood into pieces that are assembled into furniture. Steelcase recently invested approximately $17 million into 6 new Computer Numerical Control (CNC) machines at the wood plant that are highly connected with their IIoT database, constantly collecting data. Their IIoT database houses an exorbitant amount of data, with millions of entries and hundreds of tables. Our project focused on using this influx of data in new ways, with the main focus of our project being on preventative maintenance at Steelcase.
Project Objective
Steelcase’s maintenance system focuses on reactive maintenance as opposed to proactive maintenance. Due to the average cost of reactive maintenance being 3.5x that of proactive maintenance (approx. $70 versus $20), Steelcase spends a large amount annually on maintenance. Our team saw three opportunities to help shift the focus of this system from being reactive to being proactive. Through enhanced visibility of machine health, incorporating predictive analytics, and enhancing Steelcase’s IIoT knowledge base, our team believes Steelcase can immediately begin to convert breakdowns that require reactive maintenance into proactive maintenance tasks to save money and decrease downtime during production hours.
Design Strategy
Our team worked through various design aspects during the course of this project. In specific, we did in-depth research into Steelcase’s IIoT system and identified unused aspects such as alarms and warnings coming off the machines. We then cleaned these alarms and warnings to be converted into easy-to-use insights for Steelcase’s maintenance manager on the maintenance monitoring dashboard. We also did testing on our Poisson model to find which alarms/warnings were most indicative of incoming machine breakdowns. Lastly, during our deep dive into the IIoT system we met with the distributor of the CNC’s to learn in detail about these machines and their health as tied to alarms and warnings to provide learning resources for Steelcase in the future.
Deliverables
Our team focused on three different deliverables as we attacked each of the three project opportunities. In order to enhance visibility of machine health, we created a maintenance monitoring dashboard in Tableau that allows the maintenance manager to have insights such as an increase in the amount of an alarm that signifies poor machine health, as well as other metrics such as increased downtime and details on machine tool lifetimes. We also created a Poisson model to help forecast machine breakdowns as Steelcase works to incorporate predictive analytics into their maintenance process. Lastly, our team compiled all of our research, interviews, and investigation on the IIoT system at Steelcase into an insights report to help enhance the knowledge Steelcase holds as they continue to develop the system in the future. All three of these deliverables will provide value as Steelcase works on shifting to a more proactive maintenance system through an increased understanding of machine health and greater insights into the current state of a machine.
Value and Impact
By converting a breakdown work order to preventative maintenance, we see an average reduction in cost of 70.50%. With a total breakdown cost of $26,721.02 across one machine and $160,326.12 extrapolated across all 6 CNC machines, we can then identify the associated cost savings that would occur from converting these breakdown work orders into preventative maintenance work orders using our maintenance monitoring dashboard. This yields a range of savings from approximately $22.5k-$45k annually. With the possibility in the future of Steelcase integrating this maintenance monitoring into all 51 IIoT assets in Grand Rapids, in turn projected total savings would increase to a range of $190k-$384k annually.
Although we do not possess accurate figures for the average cost of machine downtime and damaged material, there is significant value present in these metrics that will expand beyond the value derived from converting breakdown work orders to preventative maintenance through a reduction in unplanned downtime that in the past has resulted in a loss of revenue and disruption in production. To continue to provide future value and recommendations to Steelcase and allow them to continue to improve their manufacturing operations, the team has also provided a wealth of information within the insights report regarding findings such as that of the Poisson model and a documented catalog of the IIoT database to provide a roadmap on their IIoT system for the future.
Project Information
Student Team
Neal Austensen, Nicholas Broadway, Alexander Buckelew, Craig Cassidy, Luke Eckenrod, Jackson Fisher, Justin Kaplin, Matthew Mazzacano
Faculty Advisor
Faculty Evaluator
