If you’re reading this, it means you’re the type of person who sees an inventory optimisation pun in a headline and thinks, “yes, this article is for me.” For that, I applaud you.
First, let me set out our position at Peak so you’re under no illusions:
1. Inventory optimisation without data isn’t inventory optimisation. It’s moving things around a bit and maybe having a clear out, much like you would in a wardrobe or cupboard.
2. Inventory optimisation with data, but without artificial intelligence, is like using an electric toothbrush but never turning it on. Your teeth might end up clean, but they’ll never be as clean as they might have been, and you’ll have worked harder than you needed to for the privilege.
It will go without saying to a lot of you inventory optimisation experts out there that, when it comes to business, stock or asset inventories need to be well-managed in order to be efficient. Similarly, you’ll also be aware that keeping inventories stocked to the right levels can often be a tricky balancing act; too much stock ties up cash and can result in wastage, while too little results in shortages, delayed outputs and missed revenues.
What we’re all looking for, then, is how to manage our inventories in the best possible way – the holy grail that is inventory management perfection. The answer to this, of course, is artificial intelligence.
The isn’t exactly a secret – AI has been mentioned in just about every article about inventory management this year (give or take.) However, what these articles often don’t explain is how it actually works. Let us be the exception to the rule with this concise yet illuminating explanation of how we go about it…
In short, we track and model demand for the stock or assets in an inventory based on variables like transaction data, replenishment and inventory policies, as well as external influences. We then build forecasts for demand using a variety of methods, ranging from simple distribution fitting to more advanced forecasting methods.
These forecasts, alongside mixed integer programming, allow us to determine the optimum stocking levels for inventories across one or multiple different sites, ultimately giving our clients accurate product listings, improved efficiencies, reduced inventory costs and minimised disruption to business.
For one client in particular, a company that specialises in equipment hire, we provide optimisation across multiple sites nationwide. Based on the forecasts of demand at its different locations, we use integer programming to decide on the optimal locations for inventory items. In doing so, though, we also need to minimise costs while still satisfying demand at each depot. You can have all the data you need to do that, but you’ll need some AI help to actually make it happen.
Our recommendation to any business that deals with inventory optimisation? Turn the toothbrush on.