Predictive Workload and Operations Scheduling

A few potential opportunities that vary in technical and political complexity. I need to line up support internally before being able to commit to any of the following: A) Predicting work content We operate a number of heavy repair facilities for performing maintenance and upgrades on our railcar fleet. When a car is scheduled for maintenance and arrives at a shop, the extent of labor required is not known until the car is thoroughly inspected and an estimate is written. Miles traveled, # loads/unloads, commodity properties, customer handling, etc. all can influence the amount of repair work required. Being able to better predict maintenance would improve scheduling of rail cars and allocation at different shops. We have a rudimentary system in place based on random forests built in Python (pandas, scikit-learn, Orange) / Excel, but it needs refinement. Blue sky objective would be to build this insight into a tool that helps direct shop loading decisions (e.g., can we justify sending this car 500 miles further away to a shop with more available capacity?). B) Predictive Maintenance Our engineering group is dipping their toes in this space. They recently completed a project to predict the need to replace worn wheels in advance of exceeding a threshold which allows the railroad to perform the replacement at a high cost to us. Not sure what else our engineers might be pondering, but I'll look into it. C) Inventory Optimization We are in very high-mix / low-volume space with many of our externally sourced components and also experience variable demand and lead times. There is an opportunity for analysis of current material management costs (carrying costs, stockout cost, inventory space) and optimum management strategy.