Storing inventory is a primary function at warehouses, distribution centers, fulfillment centers – as are the associated tasks of searching items, ensuring that they are in their correct locations, and of course – counting them accurately. Inventory may be stored on the ground (‘bulk storage’) or on racks; rack-based inventory may include full pallets, broken pallets or individual cases/cartons.
Whether a beneficial cargo owner (BCO), a freight-forwarder or a third-party logistics (3PL) provider, each company that stores inventory must count it regularly and accurately – random counts as frequently as possible, and full counts every so often. The goal, of course, is to ensure that the ‘real world’ matches what’s reflected in its digital twin i.e. in the warehouse management system.
The complexity of managing inventory ranges all the way from ensuring location accuracy of full pallets stored one-deep in traditional aisles to counting every single box stored in a multi-deep configuration, that too on tall racks in very narrow aisles. Warehouse inventory counting strategies also range across a wide spectrum – from random cycle counts done daily by one or two people to year-end full cycle counts involving an army of over-time staff working 24×7 to meet fiscal audit deadlines.
The frequency of cycle counts – and to a large extent the target accuracy of such counts – is the result of the trade-off between the resources deployed for cycle counts and the business value of better inventory data. This decision is further complicated by the opportunity cost of shutdowns, the timely availability of full-time/over-time staff for the counts, the seasonality of consumer demand, and strategic corporate initiatives such as automation and robotics.
The nature of each warehousing and distribution facility has a great degree of influence on inventory management, for example:
- Air cargo facilities have a high rate of inventory turnover, and hence do not break pallets; many of them, however, may choose to simply store the pallets on the ground instead of in racks, despite the resultant drop-in space utilization.
- Facilities in/near large metros, where real estate is at a premium, migrate over time from majority bulk to majority rack storage, that too with very narrow aisles, despite the associated investment in racking and special trucks/lifts that are needed to navigate VNAs.
- Facilities owned by apparel or cosmetic companies may have a high degree of sorting, packing, fulfillment and reverse logistics operations; these require a high degree of ‘touch’ and hence remain labor-intensive as they await very sophisticated automation technologies.
- 3PL sites dedicated to ‘full pallet in, full pallet out’ use-cases are likely to be quite large in area with dozens of long aisles and tall racks; however, the frequency of their cycle counts may be less i.e. once a quarter or once a month – or they made be cross-docking operations where pallets are kept on the ground, only to be shipped out in a matter of days.
While the warehousing and logistics industry is thought of as a ‘traditional’ one, recent years have seen rapid adoption of automation and robotics by this sector. The digital transformation of global supply chains has entered warehouses and DCs as well – from highly sophisticated reach trucks and conveyor systems to fully autonomous ground robots and drones.
There are now dozens of case studies where drones – manual, semi-autonomous or fully autonomous – have been deployed indoors to search & count inventory. The feasibility and benefits of such aerial inventory counts depend to a great extent on whether items (pallets, cases, etc.) are stored one-deep and whether all the barcodes to be read are front-facing.
The reality of most warehouse inventory today is that:
- 30% to 50% of the warehouse area is typically reserved for racking.
- Within the racks, pallets are broken & stored on the ground while the shelves usually store full pallets.
- Many warehouses convert broken pallets into ‘case reserve’, usually in the form of multi-deep storage of individual cartons; this improves rack space utilization but consumes labor and equipment for put-away and cycle count tasks.
- A relatively small fraction of warehouses have one-deep case reserve storage – such a configuration consumes a fair amount of effort to maintain but pays off well if frequent cycle counts are a business priority.
Thus, the degree of adoption, and business value, of inventory automation technologies such as drones is closely tied to how inventory is stored at a particular site.
What About Broken Pallets?
Obviously, racks with one-deep full pallets, carrying front-facing barcodes, are ideal for drone-based counts – since the main task is to scan the pallet barcode and map it to its correct location. However, drones are a compelling alternative for manual counts even for broken pallets (‘partials’) if:
- Pallets are broken only when taken off the shelves and kept on the ground i.e. in the pick locations, or
- Even if partials are kept on some of the shelves, there is business value in consolidation i.e. continuously identifying locations that are partially used and consolidating broken pallets so that some locations can be freed up for full pallets, or
- Items stored in partial pallets are of relatively uniform shape, size and color so that object recognition techniques can be used to count items using video & image data captured during the drone flights.
Business metrics such as dollars of inventory per square foot and location/bin accuracy will thus drive the deployment of automated drones for cycle counts.
Continuous Improvement or Disruption?
The warehousing industry typically relies on continuous improvement initiatives (eg. better inventory receipt management, better organisation in rack storage) to steadily increase operational efficiencies, reduce labor costs and ideally boost revenue and customer satisfaction. However, recent years have seen the rapid adoption of automation and robotics by this sector. The digital transformation of the global supply chain, accelerated by e-commerce, is forcing warehouses to speed up their operations – not just order fulfillment and reverse logistics but also storage functions such as put-away, cycle counting, and inventory audits. To successfully manage this disruption, warehouse stakeholders are embracing automation – starting with technologies that are less capital intensive and can offer short payback periods.
Drone Inventory Management
Drones are turning out to be an ideal fit for scanning full – and partial – pallets stored in racks in the one-deep configuration. This is driven by:
- Full autonomy in flight, scanning & charging
- Cost-effective, off-the-shelf drone hardware
- Intelligent automation software
- Ease of integration with existing SOPs and WMS
- Adjacent benefits such as inventory search, real-time video feeds, “go to” missions
- High degree of traceability with date-wise & location-wise image archives
The larger opportunity though is when inventory stakeholders in warehouses & distribution centers are willing to redesign their inventory storage configurations to realize the full potential of autonomous drones. This includes:
- Increasing the percentage of racking versus bulk storage
- Moving from multi-deep to one-deep configuration
- Ensuring one-deep pallet and case reserve is kept in a well-organized manner with front-facing barcodes
- Leveraging drone video & image data for inventory searches, audits, and consolidation
Deploy autonomous drones for repeatable, reliable, auditable cycle counts with a payback period as short as 1 year, IRR potential of > 30% over 3 years.