Atlantic Supply Chain Blog

You Can Pry this Spreadsheet from my Cold, Dead Fingers

Your sales people are using SF.COM. Your accountants and finance team are using SAP.  Quality data is rigorously being entered into your MES system. Why, then, are your planners and schedulers still using spreadsheets? In fact, they have told you they will only give them up if you pry them from their cold, dead fingers. Wasn’t your ERP implementation supposed to get rid of Excel schedules?

Yes, the ERP is used to model the planned orders and production orders, and your schedulers religiously create, monitor and modify them. Perhaps you should ignore the spreadsheet they are transferring the schedule from. Or perhaps it is time to acknowledge that the classic scheduling problem, despite all claims to the contrary, is still largely managed by tireless schedulers working in home grown Excel spreadsheets.

Fundamental Problems and the ERP Gap

Schedulers are concerned about three fundamental supply chain forces that they balance daily to do their jobs:

  1. Demand
  2. Production Capacity
  3. Inventory

99% of all home grown spreadsheets are used to determine feasible ways to meet demand, considering available inventory, and available production capacity. Basically, schedulers need a single tool that allows them to view demand, inventory and capacity simultaneously. Typically, the ERP does a reasonable job of tabulating inventory, capturing customer sales orders, and even forecasting market demand.  But ERP systems rarely, if ever, put demand and inventory in the context of finite production capacity. In short, schedulers use spreadsheets to perform finite capacity scheduling.

Many production environments include additional challenging constraints, like sequence dependent changeovers, shared equipment, constrained manpower and tank farms and silos. The ERP models none of these. Without their spreadsheets, a scheduler’s view of the plant would be so high-level that literally none of their plans would be feasible.

But Our Scheduler is an Expert

Some define expertise as the art of being indispensable. If so, then the typical plant scheduler is an expert, so much so that many plant managers shudder when their scheduler goes on vacation. Thus, the scheduler becomes the traffic cop of many plants, deciding which customer orders can be accepted, responding to production upsets, and finding feasible ways to just get the work done.  The key word in the paragraph above is “feasible”. Scheduling is a highly constrained problem, with lots of optionality.  The typical scheduler is working in a short time window.  They must get updates from the plant floor (you can’t trust that SAP inventory because it is 4 hours late), identify infeasibilities, and rapidly make changes to adapt to the current conditions. Finding a workable solution is a victory. Finding optimal solutions is a pipe dream.

In addition, the plant scheduler is typically the only person in the plant who really understands their spreadsheets. Their vacation-time pinch hitter barely survives the week. Since spreadsheets are single-user, nobody else has full visibility into the options among which the plant scheduler is choosing. Even worse, nobody else can help with this analysis and decision making.  The plant scheduler is on their own.

Nuts and Bolts

At the end of the day, planning and scheduling projects remain as relevant in 2017 as they were in the 1990’s, when the ERP revolution promised to solve the problem once and for all, but did not. The challenge is making planning and scheduling easy enough, full proof enough, and cheap enough to take on. Planning and Scheduling software has come and gone, but the formidable challenges of solving the problem remain the same:

  • Planning and Scheduling Processes are Data Hogs. Schedulers need accurate production rates, inventories, bills-of-material, sales orders, changeover costs, changeover times, and much more.
  • Planning and Scheduling Must Integrate with the ERP. The ERP has much of the data that makes scheduling a “data hog”. The challenge is getting at it, cleaning it up, and keeping current. ERP integration is a must, both to obtain key data and to synchronize planned and process orders.
  • Planning and Scheduler are Algorithmically Difficult.  No pure optimization technology exists that can find truly optimal solutions to scheduling problems.  The solution set must have sufficient algorithmic depth and power to take on a wide variety of objectives and constraints.
  • It Has to Be Fast and Easy. The whole point is to enable schedulers to test out various scenarios, play what-if games, and test ideas.
  • It Has to Be Collaborative. Many enterprises and plants are sufficiently complex to require multiple schedulers. Even in single scheduler environments, it makes sense to allow teams to simultaneously review and edit schedules.

Simply speaking, we believe that no system has existed, until now, that adequately managed the key blocking and tackling referenced above. Atlantic Supply Chain was designed to be that system.

With Atlantic Supply Chain, we get back to the basics. We handle the nuts and bolts, we make it as easy as possible, and we do it all in a cloud-based, multi-user system. Atlantic is the solution that schedulers will finally ditch their spreadsheets for.

admin's picture

About the Author

About: 

Atlantic Decision Sciences is committed to providing Supply Chain Optimization in the Cloud

Comments

User login

Forgot?