Workers in supply chain operations are increasingly using voice commands and responses to get picking, packing and replenishment tasks completed.
This week, fast food chain Burger King caused quite a stir when it released a TV ad that purposely set off any devices in viewers’ homes based on Google’s Google Assistant artificial intelligence (AI) technology.
By using the trigger phrase, ‘OK Google’, the ad instantly put Google Home devices and Android phones on alert. By following up with the question, “What is the Whopper Burger?”, it sent those devices scouring the Internet for a list of ingredients, which they then proceeded to read out loud.
Executives at Google, it seems, were not impressed. The company quickly issued a server-side update to specifically disable Burger King’s recording – but at the very least, this was a good showcase of how voice commands are becoming part and parcel of how people interact with IoT-connected devices.
Commands and responses
In the supply chain hubs of retailers and logistics companies, that shift has been underway for quite some time, according to Chris Heslop, senior marketing manager at Honeywell.
“The key drivers for voice activation in the warehouse come back to the fact that you need your warehouse workers to operate as efficiently as possible – and the fundamentals of voice lie in the proposition of being able to work hands-free and eyes-free,” he says.
Paper lists and mobile devices don’t help workers with that, Heslop adds. “You’re distracted. You’re encumbered. You need to put paper lists or scanners down in order to free up your hands. You’re not as productive as you might be and maybe you’re not as accurate.”
Working with voice makes warehouse workers 20 percent to 25 percent more productive and sees them achieving accuracy rates of over 99 percent, he claims.
That’s where Honeywell’s Vocollect technology comes in: workers wear a specialised headset that filters out the external noises of warehouse operations and enables them to receive simple instructions through the earphones. They then speak into the headset’s microphone to confirm that a task has been completed.
So, for example, they may be sent an instruction, ‘Go to Aisle 7’, from the back-end warehouse management system (WMS). When they arrive they, they confirm: ‘Aisle 7’. They might then be directed by a command to go to Slot 42 in that aisle, where they read back a three-digit check character on the slot to confirm their location. The command then comes to pick two items, which they then confirm that they’ve done.
At every stage in the operation, the worker is only receiving simple commands and confirming their actions using a limited range of vocabulary, Heslop explains – typically no more than 25 set phrases, including ‘yes’ and ‘no’, the numbers one to nine, and ‘repeat’ when they don’t catch a command the first time.
Getting workers up to speed
That’s extremely useful, he says, for getting inexperienced pickers on board during peak periods when extra seasonal workers are needed. It’s also a big help to companies that use overseas labour in their warehouse operations, since the system can deliver and receive instructions in their native languages – Polish, Lithuanian or Portuguese, for example.
But regardless of their language skills, at a time when customers increasingly expect next-day or even same-day delivery, a process that a warehouse manager has described to Heslop as “trying to outrun lava”, voice is a good way to get all workers completing tasks faster.
Honeywell Vocollect is used by a wide range of companies in their supply chain operations, including supermarkets Morrisons and Waitrose, music publisher Faber Music and bicycle company Raleigh. Many companies begin with picking operations, says Heslop, but later move on to other supply chain processes, including replenishment and packing.
A view of the future
JDA Software offers a WMS that already integrates with Honeywell Vocollect, but in time, warehouse workers could also be interacting more directly using voice with wearables, robots, machinery and other IoT-connected devices, according to Suresh Acharya, Head of JDA Labs, the company’s R&D arm.
He and his team have experimented with Google Glass and Microsoft HoloLens headsets, for example, as well as Softbank’s Pepper robot to explore the potential of voice recognition in supply chain operations.
Cognitive robots like Pepper, Acharya believes, could take over many of the physical tasks currently performed by humans – moving around the warehouse, picking items, packing them and so on – but in many cases, they’ll still be controlled via voice commands from human counterparts.
Over time, this will open the door to much wider data collection and, from there, greater customer insight, he says. “Voice is part of a wider picture that also involves machine learning. So the data that can be derived from voice commands and responses might subsequently be analyzed to help spot trends around customer demand, for example.”
Voice, it seems, is already making its presence felt in the supply chain. But over time, as we all get more used to ‘talking’ to Siri, or Google, or Amazon’s Alexa at home, it will increasingly be the preferred method of interaction for those working in warehouses and distribution hubs, too.