FUTURE MANUFACTURING
Efficient information transfer via intelligent speech recognition
Visual Inspection could be a good entry into the world of AI in manufacturing . Another , no less interesting starting point is the use of intelligent speech recognition , also called Voice Assisted Manufacturing . Tobias Lehrer-Ruhland , a consultant at GFT and expert on the topic , explains how the AI tool works .
Source : GFT Technologies SE
What are the benefits of speech recognition in an industrial environment ? Lehrer-Ruhland : Speech recognition is a classic AI application that we all know from our mobile phones or other domestic appliances we talk to instead of manually operating . Having your hands free and not having to walk long distances is even more important in the work environment than at home . In manufacturing or maintenance processes , there are lots of details that need to be written down or typed into a PC . Voice Assisted Manufacturing lets you record , assign , classify and replay this information via speech input so that your hands are free for the actual process . Work instructions , operating manuals and the like can also be made available on the move and exactly when needed .
Where does artificial intelligence come into it ? Lehrer-Ruhland : First of all , AI is in the underlying model that understands what you ’ re saying and validates the text , in other words the speech-to-text aspect . Secondly , it ’ s about mapping the content with the expert systems used by the respective company , i . e . assigning what is
Tobias Lehrer-Ruhland , Consultant at GFT
said to very specific classification systems and nomenclatures . Abbreviations , synonyms and connections have to be trained specifically so that the model understands everything in practice and , for example , correctly fills out forms or accurately retrieves information from huge , unstructured systems .
Why should people use the tool ? Lehrer-Ruhland : Language is the most efficient way to convey information , because people speak faster than they write . By directly entering information into the target system , we can avoid media discon- tinuity – and thus also possible sources of error . The system , which has been enhanced by our experts can be used by all users in their respective first languages . This is a major benefit in view of the growing shortage of skilled workers . And finally , Voice Assisted Manufacturing is designed to be self-learning – so it improves with every user interaction . l
Anke Roser Head of Marketing & Communications Germany GFT Technologies SE
23