Taking Voice To The Next Level

What if your warehouse workers could complete every operation in your facility with just one device? That is the goal of the food industry after the successful results it achieved in utilizing voice technology in picking/selecting operations. Costs are now within reach and companies are looking toward integrating voice with other technologies for use in warehouse operations including put-away, loading and replenishment, receiving, and loading pallets onto trailers.

Integrating various technologies into one device offers powerful capabilities, totally enabling workers to use the most appropriate technology for the task at hand. "We have standard product out there now addressing many applications including replenishment, put-away, and bulk transfers," says Larry Sweeney, vice president of product management for Pittsburgh-based Vocollect Inc. As equipment ages, warehouse managers plan to replace those systems with voice technologies.

Multi-modal is the new frontier in to-day's highly engineered environments. "Customers want to be able to leverage the same device or terminal they're using in other parts of the warehouse for voice applications as well," says Jeff Slevin, COO, at Lucas Systems in Sewickley, PA.

Lucas Systems delivered its first voice solution on multi-modal terminals nearly three years ago and currently has several dozen customers under contract to do the same, ranging from single-site operations to multi-site rollouts including grocery chains and foodservice distributors. Jennifer is Lucas Systems' voice-based warehouse logistics solution, which uses recorded speech in order to sound like a real person.

Customers no longer have to buy one device that runs voice applications and another device that does RF scanning functions for other parts of the warehouse that may not lend themselves to a voice application, continues Slevin.

With prices well within reach, the industry is experiencing a wave of adoption. "Software is the key to success with voice and, in particular, open software that runs on a service-oriented architecture (SOA)," notes Steve Gerard, vice president of international sales and marketing for Voxware, Inc., Lawrenceville, NJ. Another factor contributing to decreasing costs is that large manufacturers are producing audio-ID devices. "There's more hardware choice, so prices come down.

onfigurable software products, based on open standards instead of proprietary products, also help bring down costs."

A general rule of thumb some companies use to determine cost is to calculate the total cost per worker for implementing voice. "Some pioneering companies paid about $7,000 or $8,000 per worker to install voice—but that price is now down to about $3,000 per worker," reports Gerard.

Multi-Modal: The New Frontier

Multi-modal applications are still the new frontier, says Gerard. "Applications using a screen and keyboard are starting to be envisioned." Currently a combination of voice and scanning is being used in selection as workers pick to totes. "Workers scan the tote license plate and spend the next several minutes filling up that tote by picking to voice. This is a productive combination because usually the license plates are very long numbers, which are most effectively logged by scanning."

Slevin at Lucas notes before multi-modal terminals were enabled to run voice applications, they were limited to RF scanning processes. "But as the voice industry has grown, hardware manufacturers like Symbol, Intermec, and LXE—historically in the RF scanning arena—have enabled their devices to run voice applications. Now customers can use an enterprise-wide device with these voice-enabled multi-modal devices, with the voice terminal that supports scanning and provides a screen and keypad."

Combining voice and RFID produces powerful opportunities in the warehouse, notes Sweeney at Vocollect. "Voice is perfect for directing workers and leading them through a process, while RFID verifies that what the worker is doing is correct." For example, in case selection an RFID tag on a case communicates its lot number to Talkman on a worker's belt. "If that lot number is expired, the worker is alerted and the case is put aside and the worker chooses another case."

Lucas Systems is examining this powerful combination, integrating small RFID readers with voice. "RFID readers can read a particular item or location accurately, while voice can direct workers where to go and what to do and enable them to manage exceptions that occur on the floor," notes Lucas' Slevin.

Lucas has delivered multi-modal applications for numerous DC functions including receiving. "Workers may benefit from a screen to be able to see what's coming in and to verify receipt of materials, as well as an imager to take pictures of damaged materials to facilitate claims processing. Some customers use voice and RF for receiving, but in some cases it may make better sense to use the RF terminal without voice. In these cases you are not limited to having to purchase one type of hardware for voice and another device for an RF application."

Put-to-store and cycle-counting applications offer additional opportunities, notes Sweeney. The put-to-store application is used when a warehouse receives––but doesn't store––product. "Dock workers set up temporary locations for particular stores on the dock. As full pallets come in, they're taken to the staging area and broken down into store loads. Workers use a Vocollect Talkman wearable computer as they build pallets for each store and then they load pallets onto the truck. It's really a reverse picking operation."

Interweaving tasks eliminates the need for a dedicated workforce doing inventory counts during off-shifts. "Inventory applications like cycle counting using Vocollect Voice is growing in popularity as workers combine cycle-counting with picking activities," reports Sweeney.

For example, the WMS sends a worker to a location where he is instructed to first count the items because the quantity in that location is low. "Vocollect Voice prompts the worker to count the items and report the quantity back to the WMS, which does an update as a cycle count, and then instructs the worker to pick a certain quantity," explains Sweeney.

Multi-modal applications like picking and scanning or auditing orders are also growing in interest. "Tracking food products along the supply chain is becoming very important and capturing long lot numbers is most accurately done by scanning," says Gerard at Voxware. Workers use a wearable device with a scanner built into the voice unit so they can efficiently switch from a scanning operation back to voice.

Some companies might need to verify products being picked. "You might have high-value items like expensive cases of champagne where workers need to scan the UPC code they are picking in order to have an additional level of verification," Gerard notes. "In situations like these, you give up some productivity, however."

Billing stores accurately used to be an error-prone and lengthy process. Sweeney notes that weights of meat products heading out of a DC need to be recorded. "This is because a 15-pound box of filet mignon in reality might weigh a bit over or under and the DC must bill the store accurately." In the past, order selectors would have to write down every weight in one location, and then proceed to the next location. They would then pass along those notes to a person who would enter the information into a computer. "Vocollect Voice prompts the worker for the weight of every case and the system does an on-the-spot error check to see if the box is within tolerance and passes that data electronically to the billing system, so all the interim steps are eliminated."

On outbound side, Vo-collect Voice tracks where pallets are loaded in trucks. "Not only do the loaders tell the system they are putting a load on the truck, but they can identify where on the truck that pallet is going," explains Sweeney. "Drivers know where every pallet is, saving unloading time at destinations."

Heads Up!

Voice technology allows workers to operate hands-free and eyes-free, so they can focus more completely on the task at hand. The old way was tailor-made for all sorts of mistakes. Workers would have to walk to a control desk to get a long sheet of paper listing everything they were to pick. They carried the list and a pencil around, crossing off products picked as they worked their way through the warehouse. Any mistakes meant the process would have to be redone.

Voice improves worker productivity, usually cutting labor costs significantly, especially in the picking area. "Labor costs here are the number-one cost for a DC in the food industry," notes Slevin at Lucas Systems.

The biggest benefits to voice are in-creased productivity and near-perfect accuracy. Gerard reports companies increase productivity by around 15 percent to 40 percent, depending on the process, and one food company's error rate is about one mistake in every 6,000 picks made.

Slevin at Lucas adds a customer had an error rate of one error per 1,000 cases, but that changed to one error per 8,000 cases only a few weeks after implementation.

Voice is an excellent training device, without requiring a supervisor to follow new workers around the warehouse. "Workers grow in confidence and productivity a lot faster," adds Gerard. "This is great for food companies that have to deal with training large seasonal workforces, who can now get up to speed very quickly."

Worker confidence is an ancillary benefit most companies did not anticipate. "Workers describe using voice similar to having a supervisor helping them do their jobs by providing information they need," Slevin says. "And managers have access to productivity information with visibility into their operations, which enables them to make more real-time decisions in allocating their workforce across the DC."

Incentive plans are a way companies reward their workers for higher productivity and accuracy. "Before voice, workers would try to go faster but they made more errors," says Sweeney at Vocollect. "Now they can go faster and still be accurate, so they look at voice as a way to make more money. And companies report worker satisfaction is up."

Although companies generally choose voice to reduce operating costs across the board, some choose voice because it's a way to increase throughput without having to invest in expanding a facility or building a new one, notes Sweeney. "We hear this much of the time."

Safety is yet another byproduct. "Our customers tell us workers' compensation claims drop once voice is deployed," continues Sweeney. "This is because workers are not looking down at a piece of paper or carrying handheld devices while operating equipment like a pallet jack—so they can work with their heads up and they're more aware of their surroundings."

Voice Grows Up

In the early days of voice, the focus was on the voice hardware unit, and many companies found they were stuck with inflexible proprietary software architectures out-of-step with IT standards adopted by their IT departments over the last several years.

Open voice software solutions are popular as customers search for voice solutions that run on a variety of hardware designed to perform multiple capabilities. Customers don't have to get locked into one vendor for their software and hardware requirements, which means they can shop based on price.

Voxware offers an SOA off-the-shelf software product. "It's fully configurable, so you create the voice solution by assembling the building blocks instead of writing programs," explains Gerard, adding this is a relatively new development in voice solutions. "This allows customers to have more control in having a tool set they can use to make changes they need to make. They are also implemented a lot faster at less cost."

Integrating voice directly with a WMS solution is nothing new, notes Slevin at Lucas. "The real issue is choosing the best approach to add voice to your current system, and we've seen this is a hot topic. It will differ for each warehouse in terms of whether the best solution is a direct interface or whether it's a middle-ware or three-tier solution where the voice application runs on a separate server and integrates in real time with the WMS."

VoiceLink 3.0, recently launched, is an SOA-based middleware software that integrates easily into a company's current IT enterprise. "It's a Web-server-based approach," explains Sweeney.

"Voice is no longer new and people need to know the software has grown up and costs have come down," Gerard says.

"This is opening up the way for many new uses of voice such as multi-modal. From our standpoint, voice-only hardware devices of the past are no longer a factor in making a decision. The real decision should be focused on having open voice software.

 

 

Loading