As you’ve learned about artificial intelligence (AI), you may have decided that it’s a tool for writing social media posts and answering quick research questions. But the truth is that AI has already found a handful of “killer” applications in the factory.
In our blog post about AI for better maintenance, repair, and operations (MRO) data quality, we showed how AI is already being used to improve the quality of MRO data—but it can do much more for the factory.
Improving data is only the first step toward a number of more advanced AI applications in the factory. Applications that will improve your factory’s efficiency, reduce capital expenditure, and give you an edge over your competition.
In this blog, we’ll explore one of these advanced applications for AI in the factory: optimizing MRO spare parts inventory.
We’ll look at:
The spare parts inventory problem
Inventory is an especially complex topic in the MRO space. As you may know, MRO teams can often wait weeks or months between ordering and receiving factory parts from a supplier. The industry often resorts to the simplest solution to address this issue—keeping spare parts inventory on hand.
This approach ensures that no factory has to wait for parts when a machine breaks down. However, it introduces a new problem: MRO professionals must now figure out the right amount of each spare part to keep in inventory.
This problem is very sticky. If the MRO professionals get the number wrong, it can have significant consequences for MRO teams and the business as a whole.
When MRO teams maintain a “lean” inventory with fewer parts, they can reduce expenses, but it also increases the risk of not having the necessary parts on hand when a machine breaks down. This leads to downtime, lost productivity, and substantial revenue loss for the factory.
As a result, MRO professionals often stock more spare parts than they think the factory needs—just to be on the safe side. Unfortunately, this ties up capital in inventory that sits unused on the shelf, preventing the factory from allocating those funds to higher-ROI applications. Additionally, it requires more storage space, increasing the factory’s footprint and resulting in both capital and material waste.
“Factories buy spare parts inventory because they think they need it, and then they never need it. It’s common that 50% of the parts at any factory never move. So, what happens when they buy a new machine? They don’t need the part anymore, so they scrap it. It’s a capital and ecological catastrophe.”
says SPARETECH CEO Martin Weber.
How factories optimize spare parts inventory today
Too little inventory leads to issues, while too much creates different challenges. According to Jack Reinke, SPARETECH Senior Account Executive, the solution lies in inventory optimization.
“You want a balance: carry as little as possible, but don’t run so lean that it has a negative impact on your operations. You want to have exactly what you need and nothing more. That’s an optimized inventory in the perfect world.”
The best and most common method currently available is to set a minimum and maximum number for each spare part in the inventory, which is also known as min/max.
Set minimum numbers at a requisition point or reorder point.
For example, let’s say a motor burns out in three months, but it will take six months to receive a replacement. You might set your reorder point to trigger an alert when you have two motors left. This gives you three months until the motor in use burns out, plus an additional six months until the replacement motors in inventory are used up. This creates a buffer of nine months—six months longer than the delivery time required.
Set maximum numbers according to budgetary constraints. In other words, don’t buy more parts than you have the budget for.
Why min/max doesn’t work
In theory, this kind of min/max process is simple. In real-world MRO processes, it doesn’t work for several reasons.
For one, most MRO professionals don’t have accurate and reliable data on the spare parts in their inventory.
When they open an ERP system—if their factory even uses one—they will find inaccurate, poorly formatted, and difficult-to-understand records. If they decide to inspect the inventory in person, they’ll likely find spare parts stacked on top of one another without barcodes or records in the system.
This means MRO professionals can’t be certain about what’s in their inventory. They also can’t track when they reach the minimum or maximum number of parts, making it impossible to implement effective min/max optimization.
Even if that issue were resolved, MRO professionals still lack data on how long each part typically lasts before it breaks. As you might remember, we used this data point earlier in the blog to calculate the absolute minimum number of spare parts to keep in inventory. Without this data point, min/max optimization is still impossible.
"How can anyone set a min/max on parts that are not even in their system? Or when they don’t know how often parts break?" asks Reinke.
“That’s why optimization is not as simple as setting a minimum and a maximum. If MRO teams don’t have the data, how can they start to optimize?”
How AI can help
Fortunately, AI can help make these problems a thing of the past.
Our blog post about AI improving MRO data quality showed how AI can help MRO professionals obtain better spare parts data. The process is simple: AI cleans and enriches existing inventory records by pulling verified information from a spare parts catalog. This results in accurate, easy-to-read records that are fully compliant with relevant standards and governance rules.
We also showed how AI will make it fast and easy for MRO professionals to enter new records for all incoming inventory, helping to keep the records up to date. AI will also use image and text recognition to identify parts in a factory’s inventory for which there is no record.
In short, the blog post demonstrates how AI will ensure that MRO professionals finally have high-quality data on their spare parts inventory. This, in turn, sets the groundwork for truly effective inventory optimization.
With complete records, MRO professionals can analyze their spare parts data to extract insights. They can track how often each part is reordered and use the data to calculate the average lifespan of that part. They can track real-world delivery times to get an average shipping speed.
As such, AI will enable MRO professionals to determine a minimum and maximum for each part, enabling them to optimize their inventory effectively. This will reduce unnecessary stock and waste while also limiting risk and prioritizing uptime.
"In the past, spare parts purchasing was a safety net against machine downtimes, leading to overstocked warehouses. With the help of AI, inventories can be tracked and connected so that MRO procurement teams buy only what's truly needed based on real-world demand."
says Felix Dosch, Senior Account Executive at SPARETECH.
It’s a win/win for MRO teams and the factory overall.
How AI can take spare parts inventory optimization to the next level
But that’s not the end of AI’s power for spare parts inventory optimization.
As MRO teams gather more data about their inventory and the factory over time, they build a robust historical data set. By feeding this data to the AI, they can train the AI model to make accurate predictions about the spare parts inventory in the coming months and years.
With this glimpse into the future, MRO professionals can optimize their inventories even more effectively.
“Think of this, AI could train on supplier lead times and learn to predict how long delivery will take for any part right now. The AI could also train on historical usage data to warn MRO teams when the part in your machine right now will break down.” says Jack Reinke.
An AI model can accurately predict when parts will break and how long it will take to deliver them. Based on the most current factory data, it can advise MRO on the optimal time to order new parts. This advice is dynamic and changes in real time, making it significantly more powerful than a simple min/max optimization.
Martin Weber says that AI may also help factories under the same corporation optimize their inventories even further by pooling parts.
A corporation would do this by running “what if” scenarios with the AI, asking it to simulate different possibilities for spreading inventory across several factories. The AI could run inventory scenarios based on the historical data that each factory used for training, and you could pick the one that best fits your priorities.
The AI might suggest putting all the spare parts in a central inventory. Or it might suggest splitting them across factories, with parts x here and y there, and so on.
“In retail, they do inventory analysis day in and day out. With AI, we can bring it to MRO. That way, you don’t even have to have the parts on hand, but you can get them when you want. You don’t have to care about where they come from—whether it’s a factory within your corporation or a supplier.” says Martin Weber.
The result? MRO teams can reduce the number of unknown factors when making inventory decisions – enabling better productivity and lower expenses simultaneously.
What else can AI do?
As futuristic as this sounds, it’s not the only advanced use of artificial intelligence in MRO. Check out our other blogs to see how the factory of the future uses AI for: