Voice-activated services are popping up in search engines, cars and home stereo systems. However, such systems, which are generally limited to a single command like switching on a search engine or turning on the lights, suffer from two major drawbacks: They are incapable of managing complex tasks and they do not offer product-specific information or instructions. With this in mind, a team of researchers led by Dr. Michael Golm of Siemens Corporate Technology in Princeton, New Jersey, has developed a technology that uses off-the-shelf voice recognition software to literally speak to the needs of field service technicians.
Take, for instance, a technician working in a wind turbine’s nacelle, which can be well over 133 meters above the ground. Rather than calling someone who may not be immediately available or who may not know every detail of the particular turbine in question, the technician could access software known as a “Dialogue Manager” to verbally ask how to, for instance, open an unfamiliar turbine cover. “Our vision is to have a system that a technician can talk to as if he or she were talking to a human expert,” says Golm. The technician would just need a headset, could speak in regular language, and would receive the kinds of answers one might expect from a human expert. This would make it possible to perform complex tasks using both hands because it would not be necessary to hold a display device for instructions.