Artificial Intelligence (AI): Could it Help Your Warehouse?
David Gordon
Posted On July 31, 2023
Photo by Tara Winstead
The technology topic of the year has been, and will continue to be, AI.
In reality, much of AI is really ML (machine learning) but it’s not as “sexy” to talk about data being built into algorithms, to mine data and then deliver information, insights and/or potentially predict (or recommend) actions. The difference between AI and ML, from a theorist’s viewpoint, is that AI “machines / systems” are sentient beings, but let’s stick with the current nomenclature of AI for now.
The key, however, is thinking how today’s AI could help your business.
It’s in the early stages but rapidly evolving. This is a challenge for many distributors and manufacturers who are not used to technology changes occurring rapidly … and implementing them rapidly. It requires a different skill set for business leaders and technology leaders.
One person who knows distribution and warehousing, which is the essence of distribution and a key component of manufacturing, is Dick Friedman. Since the beginning of the year, maybe earlier, he’s been thinking of how AI will impact warehouse operations and how those who stock, pick, pack and deliver could benefit from AI … or should start thinking about. He shared his thoughts:
Artificial Intelligence (AI): Could it Help Your Warehouse?
Artificial Intelligence (AI) is not magic. From my perspective, in the context of distributors, AI is a combination of software and data to solve, and prevent, problems.
The “software” is usually a distributor’s ERP system, modified to include AI logic / formulas, and potentially interacting with CRM and e-commerce systems, and using “data”.
The data is a repository of information that can originate in the ERP database, from an AI request, perhaps from text messaging, emails, possibly enhanced via an Internet search and maybe other sources.
So what is AI? Software plus data. Simply put, AI is machine “learning.”
Let’s talk AI in the Warehouse
Here are two examples where AI could help improve accuracy and efficiency of your warehouse. Notice that AI combines data (“synthesizes”), makes a decision (“interprets”) and makes a recommendation (“projects”).
- Distributors and wholesalers want “high” warehouse accuracy, defined here as 1 minus (the number of mistakes in a week divided by the number of lines picked and packed in that week). But its almost impossible to achieve 100% accuracy because the cost would be very high, and some preventative measures would not be practical. So, distributors would like to know what level of accuracy they should try to achieve. AI might recommend a target accuracy level by using information found on the Internet or based upon an algorithmic formula the company develops based upon historical data and other acquired insights.
- Using picking as an example, for each item and pack size in the warehouse, software would search the Net, or another benchmark based upon a formulaic approacfor reported, or desired, accuracy levels. If enough data for an item and pack size is found, the software would calculate an average accuracy level and average variation from the average accuracy level. For a pre-selected product group or sub-group or vendor, if a suitable level is calculated and the variation is acceptable, AI would display the calculated average and any actual service level calculated by the ERP software, and the difference.
- Determining if a worker is as productive as desired requires standards (e.g., pieces/cartons per hour) for each job (e.g., picking) and shift. And it requires captured data for each worker, such as date, clock-times in and out, and cartons or pieces picked; all of which are captured via bar code scanning. Manually calculating those standards is time consuming, but AI might determine those standards in seconds. In this example, for each shift, AI-enabled software would search the Internet, or compare versus company / industry / peer group standards, for reported picking standards. If enough data for a shift is found, software would calculate an average picking rate and average variation from the average picking rate. If a suitable rate is calculated and the variation is acceptable, AI would display the calculated average rate—the standard.
To see that picking standards really are on the Internet you can go purchase the Warehouse Modernization & Layout Planning Guide which was developed by the US Navy (although) in 1985 for only $16 and/or download, for free, standard time data for pallet storage, binnable picking and rackable item picking. Yes, these may be dated, however, it shows that there may be resources available that can be integrated into models to help you improve your business … and NAED, or marketing groups, could commission similar research to assist their members.
Improving Operational Efficiency – What to Do Today.
AI is aways off for most distributors, so until AI becomes a reality here are a few tips for improving your warehouse accuracy.
- Organization Store items picked the most often closest to the packing area, and even where items are stored by “family” or vendor line, store the fastest moving ones closer to the front of the section.
- Receiving – If the unit of measure displayed in PO/put away data is not the same as that on the corresponding packing list, the receiver should note that discrepancy on the packing list or record it via a scanner.
- Put Away – If there is no permanently assigned storage location for a newly received item, or an alternate location is used, the person doing put away must record the selected location asap.
- Pull Down –The time to replenish picking locations from bulk/overflow is before daily picking begins, regardless of whether someone is using a printed pull-down list or displays data on a bar code scanner. Pulling down and picking at the same time leads to mistakes.
- Picking – To minimize picking time, items must be picked in a sequence that minimizes walking time. Rushing to make up for long travel distances causes mistakes.
- Packing/QC – To avoid repeating mistakes already made, an order checker should not be the same person who picked the order being checked.
- Loading – To reduce mistakes, the smaller and lighter items and packed cartons of an order could be placed on rolling shelves that are used only for staging outbound orders. Each rolling shelf can be pushed into or near the appropriate truck.
Dick Friedman helps distributors prevent warehouse mistakes that lose sales and customers; and helps determine if warehouse automation would be practical AND cost justified. Contact him at 224 723 5143 or via www.GenBusCon.com for a FREE consultation or more information.
Thoughts
The topic of AI gets one thinking of the areas where it can impact the business. Microsoft, as evidenced by its investment in OpenAI and launch of ChatGPT, is thinking much about it. They are regularly launching new functionality into its Office offering (especially in Excel and PowerPoint), and is adding much to its CoPilot offering which will bring interesting functionality to CRM systems.
NAED held one of its Futures Group sessions last week on AI, which generated a spirited discussion (and if there is a link made available to attendees, I’ll post it here.)
During the session, Steven Levy from Infor shared some of the things they are working on and then offered a tool to help distributors enhance and enrich their item data via an experimental webservice. It appears that the offer is free and is open to distributors regardless of the ERP system you have (and if you have issues with the link to the LinkedIn posting, email me and I’ll connect you with him.)
At the end of the day, every distributor has the capability to use AI. Whether you have talented individuals within your technology group or will outsource development (and connect with the myriad of applications that are becoming available), it all starts with data. The data can be customer information, sales / ordering information, product data and can be augmented by other types of information (weather?, distance?) much is dependent upon having the information, in a good structure (and AD, at last week’s HR, Finance and Technology Summit reportedly had a good presentation on data structure conducted by someone from Amazon Web Services.)
While functionally data management is an IT / digital initiative, it really needs to be something that senior management needs to understand, envision and think what it wants as deliverables. Here’s a good article in CIO Journal from Deloitte titled “Data, Analytics, and Your AI Strategy, to add to the discussion.
Getting data “right” is key to leveraging systems (and AI) to generate business outcomes that benefit customers, associates, suppliers and your top and bottom line.
3 Comments
Artificial Intelligence is a fairly terrible description, ripped from science fiction novels and applied by over eager programmers ready to join Starfleet. If the world could agree AI stood for Analytics Intelligence I think it’d be easier to understand and adopt, and the barrier to entry would feel much lower.
From that standpoint analytics intelligence is something every distributor can start applying today, most already having AI systems for years. It feels less like new technology and more like a good business practice. Simply put, paying attention to what the data tells you about your business will inform you on better inventory decisions.
Inventory management comes down to two disciplines – product handling and inventory investment. While scorecards and measures on warehouse staff capabilities are important, the consequences are lesser compared to the more critical question – what should I keep in inventory? Your investment in product is critical in the market, and with so many manufacturers and products, and variations in demand, it’s the most challenging part of your business to manage (even more than sales guys who like to go fishing!)
Inventory is the anticipation of demand, you buy it thinking a customer will buy it from you, and the smaller the period of time you have to hold the product the more profitable the sale (in principle). So analysis tools on the nature of your inventory and customers buying patterns is foundational. Some things to consider:
– how stable is demand on your high volume products? Predictability in this area allows you to maintain a just in time inventory model and your safety stock essentially matches the variability in stability. In other words not all A items should have similar safety stock levels, rather a variation coefficient should dictate where you need 4 weeks supply and where you need 8+ weeks supply.
– How many products are dependent on a single customer continuing to buy from you? Do you have an inventory agreement that you’ll keep stock as long as they keep buying, and they’ll take responsibility if their demand ends without leaving you with a shelf of products no-one else will buy?
– What’s the cost of not selling something from inventory but rather having to special order it? Do you have an incremental markup for non stock items, or are you eating costs and selling at a loss every time you bring something in?
– What should be stocked at a zone and what should be stocked at a branch, and should the branch be restocked by the zone or directly by the manufacturer? What’s the financial difference between each strategy?
– How much of your stock business can truly be served from inventory? Your fill rate is a measure of just the items you’ve flagged to keep in stock, but to you customer your fill rate is a measure of everything they buy from you. While distributors can talk about a 99% fill rate, their reality could be that only 80% of their stock sales are from inventory, leading to the question on how profitable is the last 20% of sales?
– How long do you keep an item in stock before the decision comes to return it to the manufacturer? Has any effort gone in to find a customer for those items, or are you simply waiting for the customer to call – customers who may be unaware of what you’re stocking and eager to sell off.
– More fundamentally what’s the cost to place an order with a manufacturer? Is that cost factored into your pricing model? Do you look at average order sizes, or orders below a dollar threshold? Do you repeatedly order the same product every week or two, adding unnecessary inventory receiving, counting and putting away for no real financial benefit? Should you be order 100 each time rather than just 50?
Internal analytics can answer all these questions and a review of this kind of data will reveal immediate areas where you can adjust your inventory for better performance. Your inventory is likely good for 80% of what you sell, but that last 20% can separate low profit from high profit.
Hi Mark-I’ve been in datacom distribution for 36 yrs. Your comments are spot on! I enjoyed reading your article. Thanks
[…] How will AI improve warehouse operations? […]
Leave a Reply
Trending Now
Changes at the Top - Wesco, Summit and Crescent
Lighting Controls ... The Role of the Distributor?
You may also like
