- Don't starve the bottleneck
- L = λ W ( thanks John Little)
- Manage the bullwhip ( reduce information latency)
- Variability Causes congestion
- Real problems are often NP Hard, use a heuristic or simulate the hell out of it
- Forecasts are always wrong, good forecasts are just less wrong and for a near horizon
Some of these you learn in school, some of these you learn on the job. Real world learnings that almost made the cut are below
- Grads students are cheaper to hire than undergrads or consultants and produce better work
- Real supply chain optimization is an exercise in contracts not math
- No one really cares about the best solution, any improvement is met with praise
- Performance Metrics only have to be representative, not truly meaningful, if they are truly meaningful and not deftly named they'll never get monitored
No comments:
Post a Comment