Voice technology hardware and software: have you heard the latest?
Using voice technology in the warehouse and DC can no longer be viewed as a science experiment. Here are three recent developments that have pushed its evolution and two companies that are reaping significant productivity benefits from voice.
By Maida Napolitano, Contributing Editor -- Logistics Management, 7/1/2008
The evolution of voice technology hardware over the past 20 years is nothing less than impressive. In that time, voice-directed solutions have evolved from large desktop workstations to small, rugged, wearable devices that are seamlessly connected over a wireless network.
Today’s voice software recognizes the spoken word despite noisy industrial environments and converts those words into digital data that can be processed in real time by a company’s enterprise system. Similarly, instructions from the system can be converted from digital data to verbal form, directing a worker to his next move. With these instructions relayed to headsets, workers keep their hands and eyes free to focus on their primary tasks such as picking cases for an order—while improving productivity and safety at the same time.
According to voice vendors and users alike, it’s official: Voice in the warehouse and distribution center can no longer be viewed as a science experiment. “The technology has certainly matured,” says Scott Yetter, president and CEO of voice software provider Voxware. “It’s gone beyond the innovation stage and has entered the mainstream.”
Larry Sweeney, co-founder and chief customer advocate for voice software and hardware provider Vocollect, concurs with his chief competitor. “After establishing itself in the grocery and the foodservice industry, voice has spread to other types of markets such as manufacturing, retail, and third party logistics (3PL) where voice is growing very rapidly.”
What’s driving this increased adoption of voice across these various markets? Over the next few pages we’ll dive into the three latest developments that have pushed voice along its evolutionary path. We’ll then visit with Smith Drug and U.S. Foodservice, two companies that have applied voice to solve two distinctly different problems—yet yielded similar results.
3 Key Voice Technology Developments
With a little help from our voice experts, we’ve clearly identified three developments over the past few years that have helped fuel its adoption. Two of these developments can be traced directly to the deployment of service-oriented architecture (SOA)-based software in voice technology.
Simply put, SOA is a type of modular programming structure that was created to better support the connection of different applications and the sharing of data; and this has done wonders for voice. The developments include:
1. Radio Frequency (RF) terminals can now support voice. “Previously, mobility devices, or RF terminals, such as those made by Motorola and LXE, didn’t really support voice,” explains Voxware’s Yetter. Over the past few years, his company has worked closely with these mobile device manufacturers to make sure its voice software could run on their devices. “With our open, SOA-based software, customers that wanted to adopt voice-directed solutions can now leverage their existing RF terminals without buying new devices,” he says.
Current RF users can integrate voice application software with their existing back-end host system. By adding headsets equipped with noise-cancelling microphones, workers can now pick with voice. If scanning is required, small, wearable scanners can be added, while leaving an operator’s hands free.
Using the RF terminal with voice also increases functionality, says Sweeney. “I could use the screen for inventory transactions in receiving for one hour and then the same device can be used for voice-directed picking for the next four hours.” In some systems, personalized voice files—including preferred dialogue language—can be downloaded at any time, meaning that any worker can use any terminal.
But perhaps the biggest reason driving this trend is that the price for an RF unit has dropped. “Three or four years ago, people paid $4,000 to $6,000 for their RF terminals,” says Yetter. “Today, these devices can be acquired for $1,400 to $1,500.”
In the end, Yetter believes that it’s the voice software provider’s responsibility to create “open” software that runs on as many of these hardware devices as possible. “Customers should have the choice of which wearable device to purchase based on who they have a relationship with and what might be the best unit for the job,” he adds.
2. Implementation of voice has been dramatically reduced from months to weeks. Previously, the adoption of voice into a user’s enterprise system involved months of programming and code generation in order to integrate with a user’s enterprise system. Yetter explains that with their new SOA-based voice software, they can easily map the data, the files, and the field that they need out of their customer’s system and into the format that they need to match the workflow for the voice system.
“This tool automates the integration process and gives us the ability to more easily support it,” explains Yetter.
SOA has also offered companies greater flexibility when adding features to a voice solution and radically reduced the costs for maintaining and expanding voice solution software. “SOA allows a customer to adopt or bring in a product (such as voice) and run it on their operating system and database of choice,” explains Sweeney. “For the IT department, SOA keeps costs low by maintaining one operating system and one database.”
3. Voice is now increasingly used for broken-case picks. The initial application for voice was in labor intensive, high demand, full-case picking operations typically seen in grocery fulfillment. It looks like that’s about to change. “If I had to categorize our clients over the past couple of years, it’s been greater than 50-50 for broken-case piece picking over full-case picking,” says Vocollect’s Sweeney. Companies that ship in less-than-case quantities are realizing the quick returns on voice, especially when compared to more expensive “pick-to-light” systems.
But to see it is to believe it. In the next two case studies, two companies share the trials and tribulations they encountered as they ventured into voice-directed picking.
Smith Drug: Voice As A Remedy
Based in Spartanburg, S.C., Smith Drug Co. is a wholesaler and distributor of pharmaceuticals and over-the-counter merchandise to more than 1,500 independent pharmacies, nursing homes, and hospitals across 18 states.
In business over 60 years, this privately-held company understands the challenges of supplying small drugstores and pharmacies that have to compete against much larger national chains. Randy McConnell, director of information systems, was cognizant that the key to success is to set high service standards while keeping costs low. “If your order comes in by 7 p.m., our goal is to have it at your drugstore by the next morning.”
But productivity and accuracy problems proliferated. “We were picking product units with paper and each picker could average only 450 units per hour,” recalls McConnell. With no bar code scanning and no RF terminals, pickers could not confirm picks while seventeen quality assurance stations were needed downstream to check what pickers were picking.
“It was hard to find a tote that did not have an error,” says McConnell. “Frankly, I don’t know how much merchandise we were sending out the door free of charge.” With pharmaceuticals costing from $1 to $10,000 a bottle, these picking errors quickly became cost-prohibitive. It was clear that the DC had to move away from a paper-based system and move toward a voice-directed system for their mostly broken case picks.
In 2005, the company made the transition to voice-directed picking using Vocollect’s Talkman units equipped with speaker headsets and microphones. “We were just opening up our new DC in Arkansas and we did not want to start that new operation with paper,” explains McConnell. Over a period of 30 days, McConnell and his team trained 70 pickers in Spartanburg and 25 pickers in the Arkansas facility—one picker at a time.
That approach to training, says McConnell, lead to a fairly pain-free implementation. “We did not change over our entire warehouse all at once,” he says. “We still had some pickers picking with paper side-by-side with pickers picking with voice before we gradually changed over.”
With voice, the company quickly reached new heights in productivity and accuracy. Pickers could now pick over 1,000 pieces per hour and a 99.99 percent order accuracy rate was achieved. In fact, accuracy was so high that they stopped verification of over-the-counter merchandise altogether and reduced the number of quality assurance stations from 17 to five. On top of that success, training time was reduced from weeks to mere hours. “I’ve had new employees come in, get a voice template, and start picking within a couple of hours,” reports McConnell.
He adds that the application of voice has also enabled the company to easily set up incentive programs, increasing employee morale and retention. In just over six months, says McConnell, the company achieved payback on their voice investments for both DCs. “I don’t know if I’d do anything different,” reflects McConnell, “other than do it sooner.”
U.S. Foodservice: Voice As Competitive Advantage
Running operations for U.S. Foodservice is no easy task. You have to supply quality, safety-assured food and non-food products to restaurants and cafeterias in schools, universities, healthcare institutions, and the military on a daily basis—and at the lowest costs to meet the low margin, high-demand nature that’s inherent in the industry.
But conditions in its Chicago DC were far from ideal. Pickers were constantly picking either the wrong items or the wrong quantity of items for an order.
Although U.S. Foodservice is one of the biggest foodservice distributors in the country, it is also one of many. Customers complaining of shortages began checking out the competition. Alex Olejniczak, vice president of operations, was well aware that to be competitive his company needed to stay on top of its game. These costly miss picks and shorts on trucks needed to be a thing of the past. “We knew these errors stemmed from picking with paper pick labels, leading to major issues with accuracy,” he says.
Their throughput was also being constantly challenged with a high labor turnover rate which increased training costs and ultimately decreased productivity. “We knew we needed to change,” he adds, “so we decided to use voice-directed picking for the selection of all of our outbound orders in the freezer, cooler, and dry warehouse.” Why voice? According to Olejniczak, his team wanted an economical solution that would also be ergonomic, or hands-free, with flexible WMS integration and with a quick implementation period.
It took less than 30 days to implement the Voxware system, says Olejniczak, and the results have been nothing less than dramatic. With voice, miss picks were reduced by an average of 51.7 percent. The technology also reduced shorts on trucks by 17.6 percent.
With voice-directed picking, each order completed by order pickers or selectors are tracked and recorded. “Order selectors are held accountable for the accuracy of the assignments they select daily,” he adds. Along with the boost in accuracy and accountability, the food distributor also realized gains in productivity and training time. Productivity increased by 8 percent while training time for new workers to reach a standard level of proficiency was reduced by 35 percent.
Other benefits over paper-based picking, says Olejniczak, include the ability for managers to access real-time information such as selector productivity and for customers to check on order status. Finally, by leveraging the technology provider’s multilingual capabilities, the company was now able to draw from a more diverse labor pool. “The bottom line is that we now have a happier labor force,” says Olejniczak.
View All On-Demand WebcastsVoices in the warehouse: Should you listen?
06/30/2007
