A voice recognition technology to help speed up and improve the accuracy of inspections undertaken by
Developed by Siemens Business Services, the system uses innovative voice recognition techniques so that inspectors can receive instant feedback of their findings by simply speaking into a handset.
As the inspection is performed, each value — such as brake pads — is spoken into a microphone. The data is then transmitted wirelessly to a computer.
The computer is able to recognise keywords and spoken numbers and then collates and lists data as it arrives. One of the advantages of the system is that it is completely speaker-independent, meaning that any engineer can pick it up and use it instantly.
This is unlike most other voice recognition software systems which need to be calibrated over time and learn to recognise an individual’s voice, claimed Werner Baier, senior project manager at Siemens Business Services.
‘In industrial applications, this technology has an extremely high recognition rate and can provide real-time recognition,’ he said. ‘If an engineer is inputting values for brakes into the system, he needs instant feedback as to whether these values fall outside acceptable parameters — he cannot wait that long for the system to understand.’
One of the biggest challenges for Siemens was the noise filtration element of the technology, said Baier. It uses an ambient noise filtration system as part of the software which is calibrated to filter out noises that are characteristic of an industrial setting.
Engineers at TÜV issue certificates of car roadworthiness, much like the MOT in the
‘This new system should really help speed up the overall inspection process and improve the accuracy of the data,’ he said.Baier added that TÜV has just completed trials of the system and is eager to start rolling it out.