Inside Logistics

Pick to vision

A new approach to reality in distribution centres

May 7, 2014
by Justin Doan

We live in a time where the ability to bridge our physical reality and the virtual world is making us more efficient decision makers. From the Google Glass Project to Oculus Rift’s virtual reality gaming headset, we are constantly surrounded with new ways to access information faster and smarter in order to make the right decision.

While many of these technologies are being built with consumers in mind, some may prove to be more valuable when applied to specific industries. Let’s take the Google Glass—camera-enabled, head-up display (HUD) headgear for everyday life—and put it in a food distribution centre. This new technology could help improve productivity by offering a true hands-free technology with a natural way to gain crucial data.

The evolution of picking technology
First, let’s go back in time to when picking methods were purely list-based. During that period, an operator had his “shopping list” and fulfilled it as his shift progressed. This method is low risk, low cost and gave the worker some foresight on his pallet build. Unfortunately, this approach yields a lower pick rate since it requires one hand to hold the list—resulting in a longer cycle time and low pick accuracy, not to mention the fact that a list can’t give feedback to your warehouse management system (WMS). This method is best suited for distribution centres with low volumes that cannot afford to implement new technology. As time went on, pick lists evolved into labels, which slightly improved accuracy.

With the rise of information technology, a new type of picking method emerged in 1987: the radio frequency (“RF”) pick by Shirk and Bredenbeck, which served as an automated documentation system. This new approach gave the employee a portable scanning device, initially in the form of a scan gun, to quickly scan information on the product and/or case she was picking, and to display different data on her work order. The RF pick significantly improved order accuracy. This method, however, still resulted in a low pick rate since the operator has to hold the gun in one hand or continuously walk back and forth to acquire the gun. The advent of more compact arm-mounted units eventually helped increase the productivity of these particular tools, freeing both hands to pick.

In the early 1990s, the pick-to-light method was introduced as a solution to the “cluttered hand” problem. With pick-to-light, the worker visually scans for lit-up locations that show the number of cases to be picked, picks the product, and confirms the pick by pressing a lighted button. This “hands-free” method naturally boosted pick rates and accuracy from list picking, but still contains a number of drawbacks. To implement such a system is quite expensive (about $300 per pick location), requires a reliable information system to support it, and mandates that you can’t have more than one order in each section. Given these limitations, pick-to-light tends to be a viable option mostly for unit-pick environments.

By the end of the 20th century, voice recognition technology or “pick to voice” became a functional option. This particular solution works well in distribution centres with a large number of items spread out across the warehouse because it gives employees time to confirm their pick and listen for the next location. This hands free method allows the operator to converse with the WMS and obtain key information (eg, location number, quantities, product descriptions) about his order. Implemented in the right environment, this technology provides an efficient process, with minimal errors.

The downside is that the system and hardware can be expensive. Also, in a sound intensive environment (ie, one with blowers, fans, etc), the picking process could be disrupted by noise interference, forcing repeat communications to fulfill one order. Voice processes can also become a bottleneck when used in a high pick-density situation. Further, the operator does not have visibility into the remaining cases to be picked; she is essentially picking blindly, which makes pallet building more difficult, increases the amount of cases requiring re-handling, encourages fatigue and reduces the effective pick rate.

What’s next?
What does the future hold for picking technology, and where should distribution centres strive to be? After analyzing the industry’s technology track-record, it’s clear that hands-free methods prove most successful in terms of pick rate and accuracy. In an ideal world, distributors would benefit from a new method that incorporates the accuracy of RF-picking with the hands-free, organic feel of pick-to-voice. Adding head-up display (HUD) to either technology could be enough to fulfill this tall order.

At first glance, the idea of combining HUD picking with older technologies might seem new or sexy. But considering HUD’s accuracy and efficiency benefits, it’s a simple way to realize the best of both worlds.

Consider an example of how moving from a wrist-mounted device to an HUD can improve operations: arm units only offer a few square inches of display space, often limiting the amount of information that can be shown, and resulting in cluttered screens. To get access to all of the information needed, the operator may need to scroll through multiple screens, taking up even more time. The operator also needs to move his arm to bring the display in focus; most of the time this can be done simultaneously with other actions, but in some cases this can detract from overall productivity.

With the HUD display, an operator has continual access to what is the equivalent of a 15-inch display that follows your every move, providing as much or as little information as needed. The information can be tailored to suit the individual needs of the operators. For example, an inexperienced employee may be shown more information than a seasoned employee. Saving a few seconds here and there might not seem that significant to the bottom line, but once you start multiplying those seconds by thousands of picks on a daily basis, and hundreds of employees, the monetary advantage of HUD integration becomes obvious.

HUD can equally elevate the power of voice-picking operations. HUD technology can alleviate the two main concerns about voice-picking: the time needed for the system to communicate the information to the user (it sounds like an automated phone menu) and the “blind” picking issue. The operator would have his assignment and shift performance displayed in real-time. It would show him visual cues like GPS direction for the best way to get to his next location (eg, avoiding busy aisles), staging dock information, vehicle battery levels and other useful metrics. Implementing an HUD in your operations gives employees vital information to allow them to make decisions faster, while keeping both hands free.

Beyond making the individual operator more productive, having headgear equipped with a miniature camera improves communication between team members. These cameras could be used as barcode scanners to confirm UPC or product locations, or even to report damage. It would give supervisors access to hundreds of pairs of eyes, creating a real-time feedback loop on the state of the distribution facility.

A not too far-off future
The idea of such an avant-garde project is very tempting for any tech enthusiast, but each new technology implementation has its challenges. Given the newness of HUD technologies like Google Glass, we don’t yet have all the answers or hardware required to introduce it (without hiccups) into the distribution world—but it is only a matter of time before we do. The current displays will need to be stronger to withstand industrial environments (eg, extreme temperatures, high humidity). And, as with most new gadgets, there will be a significant training period required to ensure all employees use the tools safely and correctly. Until that time, we can prepare ourselves by creating the right way to display the information and establishing standard operating procedures (SOPs) for how to integrate HUD with existing infrastructure.

There’s a plethora of questions we could debate on HUD technology. Is this really the next big thing in picking technology? When ready, will this technology deliver on its promise to revolutionize distribution centres? Will the advantages outweigh the costs? Will we be able to design a proper user interface that will be efficient for the warehouse environment? Is it even safe?

It’s simply too premature to answer many of these questions. For now, given the tremendous advantages of HUD technology that we do know, we cannot ignore the potential benefits it would bring to the distribution environment.

Justin Doan is a consultant and industrial engineer in West Monroe Partners’ Montreal office.