Voice-recognition systems offer advantages that scanning technology can’t touch

Improving warehouse picking operations: voice-recognition systems offer advantages that scanning technology can’t touch

Tom Kevan

Order picking is probably the most labor-intensive function in warehouses and distribution centers. In fact, this process accounts for nearly 50% of their labor budget. So it’s only natural that companies look for ways to automate this process so that they can increase efficiency and improve service to meet the rising expectations of their customers.

The challenge is to find the right tools to increase accuracy and improve picking productivity, and thus reduce labor costs. Take it one step further, and you have to decide which technology is most effective: bar code scanners or voice-recognition devices?

One Answer

Maines Paper & Food Services, the fifth largest food service distributor in the U.S., has answered that question. To supply the restaurants, health care facilities, and educational institutions that are its customers, Maines has built an extensive distribution system that includes six facilities. The flagship of the system is a 300,000-square-foot distribution center near Binghamton, N.Y., where 60 to 70 order selectors work each shift. In an average week, they log 400,000 picks in this facility alone.

When Maines decided to upgrade the warehouse management system in the Binghamton distribution center, it combined Manhattan Associate’s PkMS supply chain execution system with the Vocollect Talkman voice-recognition system, and then it tied it all together with an 802.11b (Wi-Fi) wireless LAN.

Maines claims that Vocollect’s Talkman system enabled it to reduce errors 25% to 33% compared with the old system. Bill Kimler, Maines’ director of systems and inventory control, says that the company is moving toward achieving an error rate of 1:9,000, which represents 99.99% accuracy. On the basis of savings related to improved performance, the company believes the voice system will pay for itself in its first year of operation.

According to the food service distributor, the benefits of the voice system were multiplied across a high-volume, labor-intensive environment. And the improved performance was a direct result of the fact that the system provided order selectors with a natural and intuitive picking system.

But not everyone agrees with the assertion that voice delivers greater accuracy. “Voice is not going to increase accuracy,” says Steve Christensen, Symbol Technologies’ principal of warehouse mobility solutions. “A voice application is best suited in an application environment where accuracy of inventory isn’t as much of a factor. In those instances, if you have problems with accuracy, which is probably a systemic problem, voice is only going to compound the problem.”

Hands Free, Eyes Free

Nonetheless, voice-recognition systems let users concentrate on the selection process, freeing them from having to focus on the tool that they are using. This simplifies the function and streamlines physical movements by eliminating the necessity of picking up or putting down an instrument, or shifting the gaze of the employee from the instrument’s display to the material being collected. All this translates into more efficient physical movement and reduced opportunities for error.

“Instead of having to reach to scan a label when picking, with voice technology a spoken check digit easily confirms the correct location,” says Aaron Miller, principal at Tompkins Associates Inc., in the white paper titled Order Picking for the 21st Century. This demonstrates a key benefit of voice because operators can focus visually on the assigned task rather than on a technological device. Voice enhances an operator’s ability to concentrate on a given assignment by eliminating excess physical movement that can sometimes be a distraction. For example, bar code locators require operators to scan the item and then key into the device the quantity picked. Looking down at the scanning device to key in the information diverts their attention from the inventory location slot. This diversion can allow them to lose track of where they are in the process. If they key in the wrong information, inventory inaccuracies result. “With voice, operators are continually being prompted, keeping their concentration and attention levels high and focused directly on the pick itself,” writes Miller.

One of the biggest advantages of a voice-directed picking system is that the order selector’s hands are free. “The hands-free function offers productivity gains over a scan-intensive application,” says Christensen.

“It’s particularly important if you’re in an environment where you’re picking a high volume of small pieces,” says Sherri Gansen, principal marketing manager for voice-recognition products at Intermec Technologies.

Voice’s Man-Machine Interface

Workload assignments for the order selectors are downloaded to the voice devices from the warehouse management system. The system prioritizes the order picking process by grouping orders for efficient picking. The warehouse system can download this information to the order selector in real time, or it can place the information in a batch mode.

“Voice systems allow operators to communicate directly with the warehouse management system, labor management system, or proprietary host system to pick orders quickly and efficiently without using handheld devices or paper to record picks,” says Miller.

Also, voice-recognition systems provide a simple and intuitive interface for the order selectors. “With a voice solution, there is no interpretation of the requested action the system asks the user to undertake,” says Tim Eusterman, vice president for marketing and business development for Vocollect. “They simply receive a verbal command, which everyone can easily follow, and then they respond to the system by speaking their status or actions. What makes this unique is that users have no responsibility to make a judgment on the requested action. They just follow the instructions in a very human way, by listening and responding with spoken word. If they did not understand the instruction, they can simply ask for it to be repeated.”

Other technologies require some form of interpretation of the work process, the instruction presented, or both, says Eusterman. Whether directed by a printed page, labels on a sheet, or a display on a bar code scanner, users have to interpret the information presented to them. “This is where these systems begin to lose their effectiveness versus voice,” Eusterman says. “Both in terms of accuracy and productivity, a system that relies on the user to input his or her judgment during the process will not be as effective as a voice-based solution.”

The Bar Code Alternative

Bar code scanning can enable data accuracy rates as high as 99%, which makes this technology superior to manual data collection. But within the context of picking operations, bar code scanning has its drawbacks.

“Bar code read rates can be affected by environmental conditions, lighting, dirt/smudges, and print quality,” says Miller. “For scanning to be effective, all labels must have bar codes that comply with standard size and format specifications, and the label must be in good, readable condition. To read codes, a direct line of sight within a specified read range is also required for successful reads. Often, as an operation grows, the number of users connecting through a single access point can cause response delays to users and degrade productivity.”

Voice’s Payoff

Associated Wholesale Grocers (AWG) is one of the largest grocery wholesalers in the U.S., with a 1 million-square-foot facility in Kansas City, Kan. To improve performance of picking operations in its distribution center, AWG replaced the scanning technology it was using with a voice technology-based solution.

“Since the implementation, AWG has seen significant improvements in productivity in all areas of the distribution center,” writes Miller of Tompkins Associates. “The greatest improvements in productivity came over existing scanning operations in produce, meat, and dairy areas ranging from a low of 8% improvement in produce to a high of 15% improvement in dairy.”

For grocery distributors in particular, voice solutions seem to do better than scanning technology in cold storage areas.

“Speech technology is effective in cold storage areas, where temperatures can get down to 30 degrees below zero,” says Intermec’s Gansen.

“Voice does not require the need to remove gloves to enter information, to wipe frost off screens, or to remove devices frequently to thaw the liquid crystal displays,” adds Miller.

Improved productivity is perceived across the board. “Generally, there is higher productivity in a voice picking operation because your hands and your eyes are free,” says Symbol’s Christensen. “So high productivity is generally achieved, and you can expect that to be 10% to 15% more productive than a scanning-based operation.”

According to the Tompkins report, AWG also cut costs by nearly $250,000 annually through savings on bar code labels. Additionally, AWG saw a significant reduction in training time for order selectors. Training benefits are often overlooked when companies evaluate the benefits of voice systems. “The labor and training benefits of voice can be much greater [than scanning],” says Christensen. “[The labor force] can become more comfortable with the application.”

“One thing that jumps out at our customers is the decrease they see in new employee training times and the overall satisfaction of the workforce,” says Eusterman.

On another front, order picking errors frequently occur because of mispicks, and they can contribute to an operation’s overall costs through returns. After implementing voice in its Kansas City facility, AWG saw a rise in order accuracy.

“For companies like AWG, who have high volumes and thousands of SKUs, returns represent a significant cost,” says Miller. “At AWG Kansas City, the 50% returns reduction after implementing voice resulted in nearly $1.3 million in savings in the first year.”

Company Information

Intermec Technologies Corp.

Everett, Wash.



Symbol Technologies Inc.

Holtsville, N.Y.



Tompkins Associates Inc.

Raleigh, N.C.



Vocollect Inc

Pittsburgh, Pa.



Voxware Inc.

Lawrenceville, N.J.



The Benefits of Speech

Features Benefits

Hands-Free, Eyes-Free More concentration on the

Operation picking process, less on the tool

or technology, simplifying the

function and streamlining

physical movements–enhanced


Intuitive Interface No interpretation of requested

action required when verbal

clarification is immediately

available; less training required

to get new operators up and


Reduced Impact of Less affected by lighting and

Environmental Conditions extreme temperatures

Productivity Increase After Voice at AWG

From Paper to Voice

Dry 3%

Freezer 4%

From Scanning to Voice

Produce 8%

Meat 12%

Dairy 15%

Source: Tompkins Associates Inc.

Note: Table made from bar graph.

Reductions Achieved After Voice at AWG

Shortages 11%

Mispicks 25%

Returns 50%

Source: Tompkins Associates Inc.

Note: Table made from bar graph.

Web Resources

See Frontline Solutions’ exclusive online table comparing the features of products from speech-recognition

Extra! Extra

For further reading on this and related topics, see these articles, available at www.frontlinetoday.com/102003link

Associated Wholesale Grocers Replaces Scanners with Vocollect’s Voice-Direct System” March 2004

“Intermarche Optimizes Warehouse Operations with VoiceLogistics from Voxware” January 2004

“LXE, Voxware Team for Speech-Based Warehouse Systems” December 2003

COPYRIGHT 2004 Questex Media Group, Inc.

COPYRIGHT 2008 Gale, Cengage Learning