Proven financial and productivity gains have been firmly driving the supply chain logistics industry down the path of adopting Voice Directed Solutions for many years now. It is, after all, the most natural and efficient way for warehouse workers to interact with technology in order to complete their task at hand. It is the only technology that allows them to keep their eyes free to ensure they are selecting the correct product, and their hands to be free from papers or handhelds so that they can quickly and safely keep products moving.
While the choice for using a Voice Directed Solution versus handhelds or relying on paper based systems is simple, there still remains the selection of which type of Voice Directed Solution to go with – Speaker Dependent or Speaker Independent.
Simply put, inside of every device used for voice directed technology there is a voice recognizer which does exactly what it says – it recognizes the voice of the user and helps direct them to accomplish their work. How well this voice recognizer works directly affects how quickly, efficiently and accurately the worker completes their task. If a voice recognizer does not “understand” what the user is saying, then it is very much like when you are on a call to an automated line to a company having to repeat yourself multiple times or ultimately being directed to the wrong department. It costs you valuable time and prevents you from achieving what you set out to do – oftentimes forcing you to start over from the beginning. This is an example of Speaker Independent (or Untrained Voice) technology.
Speaker Dependent (or Trained Voice) recognizers are designed for a single purpose – to recognize the voice of one user and to continually adapt and learn that user’s voice as they continue to do their job. When first deploying a voice directed solution in a warehouse, each user would spend about 15-20 minutes training a set of words that are used to complete their tasks. This is a one-time minimal investment as it is paid back within 6 days. From here on out, the device continually adapts to that user, say for example one day they have a cold or their voice changes for some reason – the Speaker Dependent system will always recognize the user. It also removes any complexities for warehouses that have users with varying dialects or native languages. The Speaker Dependent system will work for all of the workers, all of the time.
Looking at it from a financial perspective, which is one of the primary reasons for selecting a Voice Directed Solution, the numbers are fairly straight forward. In a 2011 study performed by Vocollect, the world’s leader in voice technology solutions, an analysis was performed which shows that the cost of a 1% word error rate is €300 per user per year. This is based on the average time for an error recovery being 3.5 seconds. The typical case picker speaks around 6,400 words per day which would result in approximately 64 errors that need to be recovered every day. Multiply this across how many workers you have in your warehouse, and you start to see the numbers add up. More significantly are the errors that are not corrected in a Speaker Independent Solution which result in the wrong product being sent to the wrong location and the administration and handling costs to repair those mistakes.
Selecting the right Voice Directed Solution to implement is critical. There have been many stories throughout the industry even recently of workers being slowed so significantly by having to repeat or correct errors, entire systems have been turned off and the warehouses returned to their paper based systems. Unfortunately, examples such as this make people wary of Voice Directed Solutions as a whole. The best advice to anyone considering a Voice Directed Solution is to thoroughly investigate the two different types of technology side by side, not in the board room, but in their own warehouse which is filled with every day operating background noise. This is the only way a customer can satisfactorily put the Speaker Dependent versus Speaker Independent debate to bed, and any solution provider should be more than happy to accommodate such a request.
In the end, some voice providers will communicate that their solution is “good enough”. Heavey RF Group knows the importance of implementing quality solutions which bring about the largest returns for our customers globally. We don’t believe in “good enough” – we believe in getting it right the first time.