ATTN: Business and company leaders whom aspire to excel,
Meet Drew Harnish, P.E. (B.S. and M.S., industrial and systems engineering, Oklahoma University), senior project engineer, Lockheed Martin
There are tens of thousands of unique parts that are assembled to make the F-35 Joint Strike Fighter and dozens of locations across the globe that require shipments of parts. The bulk of the part consumption remains at production sites but as more jets are delivered to our customers, there is increased demand for spares and upgrades at military bases around the world. The part demand requirements and associated supply replenishments are captured in our Material Requirements Planning (MRP) system which allows analysts and buyers to see tabular snapshots of real time demand and supply for any given site at any time. As our production rate begins to ramp to triple the current rate and more F-35 bases are stood up, it has become an increasingly manual task to make decisions on supply allocations when challenges arise.
Consider this scenario: a fielded jet is grounded in Arizona due to a damaged part and the base requests a spare to be delivered ASAP. An items analyst sees this request, notices a part available in the Texas warehouse that hasn’t been allocated to an aircraft, requests a transfer, and solves the problem. The base is happy. Easy, right? Next, a production buyer receives a notification that there is not enough supply on hand to meet discrete demand in the factory. The buyer is confused because there was enough on order and there have been no scraps on the line. The reallocation results in a part shortage on the production line, increased labor cost to work around it, and increased material cost to expedite a single part which had already been purchased as part of a block buy. Program management blames production, production blames supply chain management, supply chain management blames sustainment, and everyone is really doing their best to deliver for their customer. There are countless shortage meetings to heroically solve problems just like this on a daily basis. They all stem from the same two root causes; there is no single source of total aggregate part demand or a model to account for quantity risk from unplanned part demand. This happens to a lesser extent when parts are scrapped on the production line with no safety stock in place but is exacerbated when multiple locations introduce unplanned demand into the supply chain community.
In 2014, Lockheed Martin Aeronautics Company stood up an Enterprise Integration team. This team combines the use of advanced analytics and predictive modeling capabilities with operational excellence skillsets to solve systemic problems that cut across functions. One of the first projects our team led was a multi-pronged demand and supply planning optimization effort that eventually resulted in the creation of a new organization with new capabilities. As one of five project leads, I was responsible for leading the team that developed standard business rules for solving demand/supply problems like the scenario discussed above using newly developed predictive analytics. The primary analytics model was an integrated risk-adjusted demand requirements model referred to as “Big D.” This tool aggregates all of the discrete part demand data from our MRP systems and stacks a risk-based order quantity (RBOQ) factor on top which is modeled from historical experience such as scrap, repair, unplanned field demand, and other less-than-lead-time demand. The integration of this data with the risk models allows Big D to plot two critical visuals: 1) a month-over-month demand profile from all sources including overall risk for each part which allows the user to identify parts needing stop-gaps such as safety stock before long term fixes can be put into place, and 2) a month-over-month supply profile that shows discrete and risk-adjusted line of balance (netting of supply vs. demand) to highlight areas where MRP shows healthy supply plans but the risk model indicates high potential for problems.
For three pilot areas, we analyzed the line of balance 60 days outside lead time for over 100 unique part numbers. Using the data from Big D, we developed a decision tree to triage each part based on its risk factors in order to categorize short and long term solutions for recurring issues. The methods we developed formed the basis for a new supply plan health analyst role in the company that did not previously exist. There are now multiple analysts dedicated to analyzing part issues just outside lead time to mitigate shortage risks before problems occur. This has resulted in an increase in collaboration between our production, sustainment, and supply chain teams.
The overall project team responsible for leading this enterprise transformation was over 70 people from all facets of the business and a number of consultants. The specific team discussed in this case study was made up of 16 people from Lockheed Martin and Deloitte Consulting, 5 of which were industrial engineers.”
Industrial and System Engineers provide incredible value to any organization in any industry and I am really excited to share these stories and inspire you and your company to hire ISE’s.
Blessings to you all!
President, Institute of Industrial and Systems Engineering