Wollongong researchers have studied the ultrasonic mobility aids used by blind people for new insights into the technology and how it might be used in disaster zones.
The research has produced a wealth of previously unknown data and identified the best patterns for blind people to train their aids.
University of Wollongong robotics engineer Dr Sherine Antoun led the study, which involved setting up a navigation experiment and observing how blind people used their aids to negotiate corridors.
‘‘We concluded... an appropriate scan motion appears to be: left until a wall is detected, right and down until the floor is detected, up to free space, repeat,’’ Dr Antoun said.
‘‘These insights may lead to better navigation aids, improved techniques for the use of these navigation aids and new guidelines to improve the navigability of environments for blind people.’’
Dr Antoun’s approach was different because it assumed there could be a significance in every noise – or absence of noise – produced by the aids, not just in the loudest noises indicating an object to be avoided.
‘‘Previous researchers were discarding really good data for no reason,’’ Dr Antoun said.
‘‘We now understand how some people can really work with these devices and some people can’t. Those who can have learnt they’re really not looking for loud echoes – because they’re not really interested in obstacles, they’re interested in the space they can walk through – the no-echo zone.’’
The research has implications for navigation at large, including for use finding people in extreme or hostile environments.
‘‘Big robots are very expensive and you really don’t want to go destroying too many of them, but we can equip tiny robots with the same technology and we can send many of them into a disaster zone, each bringing back a little bit of information,’’ Dr Antoun said.
‘‘We can fuse this information into a site map that can tell us, for example, that this part of the site is too narrow for a human to go through, whereas this part is not.’’
The research will be published in the December issue of IEEE Sensors Journal.