Voice Directed Warehousing:

Voice Directed Warehousing (VDW) is the implementation of speech recognition in Warehouse Management Systems. VDW was originally limited to picking, whereas now all warehouse functions (picking, receiving/put-away, replenishment, shipping) can be coordinated by voice systems.
Implementing voice systems in the warehouse has among its benefits
• Increased picking accuracy
• Increased inventory accuracy
• Increased employee productivity
• Reduced new worker training time
• Increases job satisfaction for warehouse associates
• Growing customer satisfaction
Each operator is given a voice-enabled RF device. These devices need not have screens or keypads – operators communicate with the system via headset.
Managers use the Warehouse Management System (WMS) or middleware to assign operators work – jobs such as picking, put-away, replenishment, and truck loading.
With voice picking, the voice system directs the operator to perform each pick, giving them directions to the pick location. Depending on system configuration, the operator may be prompted for a location check-digit or a container check-digit as well as a count-back. Following is an example picking dialog:
Computer: Go to building one.
Operator: Ready
Computer: Aisle five.
Operator: Ready
Computer: Bay twenty-seven.
Operator: Ready
Computer: Pick four cases of double-chocolate almond fudge.
Operator: Ready
Computer: What are the check-digits?
Operator: Four-Five-Three
Note These might be from the pallet license or location
Computer: How much remains?
Note: Optional count-back - increases accuracy and reduces speed
Operator: Seven.
…and so on, until the order is complete. The system then directs them to the location to put their pallet (onto a truck, into a staging area) and then tells them to take a new pallet and start the next pick. Unique license plates can be assigned to each picked pallet so that they can be easily located in the warehouse – and so that the voice system can direct the process of loading staged pallets onto trucks.
Apart from training users to use the voice system, the system must be trained to understand each user. This training process takes roughly 30-45 minutes - during this time, the user is asked to repeat aloud the words they will be using when they communicate with the system. The voice system learns how the particular operator speaks, and saves a template of their speech patterns. This allows voice systems to understand users with heavy accents – the system knows how each individual user pronounces each word.