If it catches on, a busybody will monitor who you speak to, for how long and who you ignore. Sensor-equipped name tags will detect when conversations take place and note the participants.
When these sensors recognise one another's identification information, the system assumes a conversation has taken place. Once recognition is made the sensors begin automatically measuring the length of conversations.
The measurement data are transmitted to a server, where they are analysed and presented in graphical form. The quantity of conversations by persons and groups is plotted on a curve, with the most-active talkers at the apex and the most-isolated persons at the base.
So, presumably, chatterboxes are to be encouraged. What if most of your conversations take place at the drinks machine, or in the kitchen. How does the quality of conversation vary as it moves from the terminal, to the toilet? They don't say. This needs further investigation, we would suggest.
A quick glance at a graph can show which persons and groups are not communicating enough during the course of work, according to Hitachi.
In one test of the system using 37 workers, Hitachi's graphs gave an indication of the exact time, during a development project, when a problem arose. This was when managers had not been interacting very much.
Hitachi plans to commercialise the system as early as next fiscal year.