The Cost of Approaching Capacity

A phone running low on storage can become frustrating well before it is completely full. Apps need room for temporary files, downloads need somewhere to go, and operating-system updates may require space beyond their final installed size. What looks like available storage is no longer necessarily usable space for the next task.
Most people recognize the problem in a phone. It is less obvious how often the same loss of operating room appears elsewhere.
A warehouse makes the problem physical.
One of the challenges in warehouse operations is minimizing the distance workers travel to retrieve products. A common approach places items frequently ordered together near one another. If workers can collect several products from the same area, each order should take less time to fulfill.
That reasoning works when considering an individual worker’s journey. But when many workers are collecting orders simultaneously, it introduces another problem.
The most convenient aisles become the busiest.
Workers and forklifts converge on the same routes, creating picking delays that shorter individual journeys do not reveal. An arrangement designed to shorten travel distances can increase the time spent waiting.
Researchers studying warehouse storage assignments have examined this precise tradeoff. In a simulation built from actual warehouse data, a method that considered both travel distance and traffic congestion outperformed one focused on keeping related products together. The improvement came from considering what happened when many workers shared the space, rather than treating each journey independently.
The warehouse hadn’t necessarily run out of room for products. It had begun running out of room for movement.
Retail stores can encounter a related problem, although the limiting resource may be less visible.
A store can have merchandise available and customers coming through the doors while becoming less effective at turning visits into purchases.
Researchers examining hourly traffic, staffing, and sales across 41 retail stores found systematic understaffing during peak periods. Their model estimated that the stores were losing potential sales because available service capacity wasn’t keeping pace with customer traffic.
An observer looking only at sales might see growth slowing despite continued customer interest. Market demand could appear to be leveling off when the store’s ability to serve that demand had become the constraint.
And retail has several capacities operating at once. Floor space limits how comfortably customers can move. Employees limit how many people can receive assistance. Checkout capacity limits how quickly shoppers can complete purchases. A store can approach one of these limits while remaining comfortably below the others.
Something similar happens with time.
Consider a manager whose calendar is almost entirely filled with meetings. Each meeting may be worthwhile, and a nearly full calendar might look like productive use of a limited resource. But unexpected questions, urgent decisions, and conversations that run long have to fit somewhere.
As the remaining time disappears, small interruptions begin displacing other commitments. The manager may be working just as hard while colleagues spend more time waiting for decisions.
The mechanisms differ. A warehouse loses time through congestion. A store loses transactions through limited service capacity. A crowded calendar leaves nowhere for interruptions. In each case, some apparent spare capacity is what keeps the rest of the system usable.
A phone can display its remaining storage. An aisle, a sales associate, or a person’s availability offers no equally clear warning. By the time one of those systems looks full, it may already have spent a long time becoming less efficient.
Sources and further reading
- Apple Support — How to check the storage on your iPhone and iPad
- Android Help — Free up storage on your Android device
- Lee, Chung, and Yoon — Two-stage storage assignment to minimize travel time and congestion for warehouse order picking operations
- Mani, Kesavan, and Swaminathan — Estimating the Impact of Understaffing on Sales and Profitability in Retail Stores
Featured image: Photo by Caleb Oquendo via Pexels.
When I was at Microsoft, we explored using a combination of the HoloLens, digital twins, and other tooling in warehouses. We prototyped a system that linked with the “Microsoft Business Intelligence” suite to advise bin layouts based on frequency of access, specifically designed to reduce overall pick time.
Our early proposals didn’t identify the issue of how if applied naively it might then later lead to picker congestion. So when researching this article I found that application particularly interesting.
That said, we did find significant value in using XR glasses to help temporary or new employees quickly find their way through an unfamiliar warehouse and to a specific bin in what is otherwise a wall of confusion for new hires. The heads-up display and hands-free computing also allowed updating their picklists and direct communication with their team lead.
This clear value is one of the reasons why HoloLens 2 was built to support wearing a hardhat (a major issue with the first generation of HoloLens that prevented our tech from being used in real warehouses). The other major issue… battery life one quarter the length of an average shift.