Inventory Optimization: Models and Simulations

Nicolas's Inventory Optimization: Models and Simulations is a game changer. I made it a goal this past year (2020) to study mathematical Inventory Optimization. Several academic books, papers, and software vendor blogs later, I received my copy of this book. Nicolas has a way of explaining mathematical concepts in a digestible way. He also goes beyond models that require such strict assumptions that they are quite unrealistic. Supply Chains are complex, you can't assume a perfect world when modeling. This book is the best I have found so far to go through more complex modeling while also trying not to scare the reader away. I recommend this book to all who are interested in improving their inventory planning - you will come away a better planner.

If you are a reasonably well-learnèd inventory manager who understands basic statistics, Excel, and Python you will adore this book. I use this book as my go-to field manual when designing inventory optimization at work. The two best books on SCM that I have ever found are this book and Vandeput's other book on Data Science for Supply Chain Forecasting. Your IQ does need to be somewhat high to get anything out of this book (if you failed statistics don't bother), but if you have the background and you're passionate about technical things you will eat this up. I highly recommend this if you're an advanced practitioner attempting to do stochastic inventory control for your firm. I highly recommend this for anyone who is in any sort of masters program in SCM too — I would say it's equivalent to roughly 8 semester hours of credits in academia. A final word: if you're into this stuff, please do yourself a favor and get on Linux. WSL on Windows + VSCode editor running WSL remote is where most people should start (if you're a business type locked into Excel). From there you can deploy your stuff on with Docker to cloud for automated compute. Mac probably works too, just stay away from native Windows for the Python ecosystem tools, that's why Microsoft put Linux into Windows so you could have the best of both worlds for prototyping things rapidly.

Good introduction to inventory management!

Nicolas in his book has a way of explaining mathematical concepts in a digestible way. We all know that Supply Chains are complex and it is not a exact science. This book shows differents inventory policies and the way to optimize them. It include too simulations and Phyton codes. Definitely an excellent book showing "what" and "how to" optimize inventory.

This is my favorite book of the three books Nicolas published thus far as of Nov 2023. I read it after finishing his other book on "Data science for supply chain forecasting", which really sparked my interest to dive deeper into these topics. Nicolas showcases how to use mathematical models to capture supply chain complexities. With each chapter, the models become more powerful and more capable to capture reality. Finally, simulations are presented as pinnacle of optimization. I had some background in Python so it was really a breeze to implement the suggested code snippets and use them as baseline to develop my own models and simulations. It was a great inspiration. Even if you are not a Python expert, you can use ChatGPT for troubleshooting and coding support! This book can help you to bring your inventory management policies to the next level, leaving outdated approaches such as ABC/XYZ + economical order quantity methods far behind and deploying modern continuous optimization through simulation. For instance, I modeled the total cost of ownership for our company: From purchase order placement through consumption of parts/delivery, factoring in dynamic transaction cost, inventory holding cost and backorder cost as function of order quantity and service level / fill rate. Using probability mass functions for demand and supply lead times based on historical and forecasted data, simulations enabled me to find optimum sets of order parameters and safety stock to operate in the proximity of the total cost minimum - fantastic! Once you grasp the concepts you realize the power this book allows you to harness.

Inventory optimization is not my background. I enjoyed reading the book and learn the fundamentals, in a step-by-step manner. Pros: - A lot of examples - Easy to decipher (advanced) topics - When and why each approach in Inventory Optimization should be used.

I continue to be impressed by Nicolas Vandeput’s books. They give you so much more than what you would find in a hypothetical academia textbook. They allow you to pivot in and out of detail and “how to” and “why” and if you didn’t understand it, here’s the “plug and play” version so you can learn by doing. It’s like getting the notes from your favourite lecturer, that show how the engine really works.

I was disappointed by this book. For context, I came to this book as someone familiar with optimization and forecasting but new to inventory. I originally bought this book to learn the nuances of quantitative inventory problems. This book excels at introducing the reader to the basics of inventory control theory. Does a great job of deriving different formulae that are commonly used in industry and discussing the different decision policies often used in practice. It provides the common language and concepts needed to understand and communicate to more experienced practitioners. The text fails at applying these methods in practice. The code examples are difficult to reproduce. They often rely on global variables that were defined in the text several chapters earlier. I often found my self rifling through past examples to understand how a particular variable or function was defined. The book makes use of local optimization methods in SciPy to identify order policies. These methods aren't guaranteed to provide a global optimal solution to optimization problems. I can't recall this being explained in the book. This can lead to sub-optimal solutions for more complex decision policies discussed in the last couple chapters of the book. It was disappointing to see no demonstration that the optimizer genuinely reached an optimal value, or discussion of the pitfalls of local optimization techniques, in any part of the book. The text has no mention of making a decision from a forecast. There are many concepts in forecasting (e.g. quantile forecasts) that naturally extend to inventory control problems and was disappointed that the text didn't demonstrate how a seasoned practitioner would use these tools.