The Sonic Pathfinder

An Electronic Travel Aid for the Vision Impaired

by Tony Heyes
Perceptual Alternatives

See the Sonic Pathfinder Training Film on YouTube
The Sonic Pathfinder

Introduction

The Sonic Pathfinder is a secondary mobility aid for use by people with a vision impairment. It is not suitable for anyone who does not have primary mobility skills. It is designed for use out-of-doors in conjunction with either a cane, a guide dog or residual vision. Sonic Pathfinder close up The use of the device must be taught by a correctly trained Mobility Instructor. Our experience has revealed that, provided the client has suitable primary mobility skills, the use of the Sonic Pathfinder may be learned in five or six training sessions spread over two or three days. The Sonic Pathfinder gives the user advanced warning of objects which lie within his or her travel path. In the absence of any such object the device switches to its lower priority function of providing information about the proximity of the shoreline. Increased safety is the main goal; however, this secondary role means that the device also provides some useful orientation information. Although designed for outdoor use, an accomplished user might well obtain some value using it in large public buildings. Use in a normal home setting is virtually impossible because of the presence of many close objects. The device will work in rain and snow; it will only fail when the transducers become saturated.

Historical Background

The Sonic Pathfinder has evolved out of the work of the Blind Mobility Research Unit at Nottingham University, England. It is the product of a research program dedicated to achieving an understanding of mobility and the information requirements of the independent blind traveler. The Sonic Pathfinder embodies many of the findings of the Nottingham program; the emphasis is on achieving a simple display. For instance, the device does not attempt to provide the user with a sonic picture of the visual world. Instead it is selective; it displays only that information which is of immediate practical interest to the moving pedestrian. The device uses an auditory display but one which has been chosen so that it does not mask the user's ability to hear, and make use of, environmental sounds. The display comprises the notes of a musical scale, which make a familiar tonal progression as the user approaches an object.

A Technical description

The Sonic Pathfinder is a head-mounted pulse-echo sonar system controlled by a microcomputer. The five ultra-sonic transducers which are mounted on the head-band comprise three receivers, one pointing left, one right and one straight ahead, and two transmitters. The two transmitters are angled so as to irradiate the user's pathway with ultra-sonic energy. Echoes from objects lying within this pathway are received by the three receiving transducers and this information is processed by the on-board computer. The output of the device, the display, is fed to one or other of the two ear pieces depending on whether the object is on the left or the right of the user or to both ear pieces if the object lies directly in the travel path.
The Sonic Pathfinder does not give information about the surface texture of the objects which produce the echo signal. This is in line with the findings of the work carried out at Nottingham which suggests that such information is of secondary importance. Another simplification is that the display is restricted to the nearest object only. The exception to this simple rule is that priority is given to objects which lie in the centre, so that, for example, the presence of a shoreline at 1 metre (3ft.) will not prevent the user being given advance warning of a centrally placed object at 2 metres (6ft.).

A Functional description

The function of the device is most easily described by cataloging a number of imaginary situations.
Sonic Pathfinder in use with a long Cane
Imagine yourself in an open space some 4 metres (12ft.) away from a wall. If you were to turn to face the wall and start walking, suddenly you would hear, in both ears, the notes of the musical scale descending in order. Each note represents a distance of approximately 0.3 metres (1ft.) of travel. If you were to stop when the tonic is reached; you would be able to reach forward and touch the wall with the outstretched hand.
To illustrate a slightly more complicated example, let us suppose we wish to explore the side beams and the priority given to objects which lie straight ahead. For this exercise one would need a long stretch of featureless shoreline which is suddenly terminated by an object which sticks out from the shore line; this might well be another wall at right angles or a buttress. Imagine following the shore line and keeping a constant parallel line. You would hear an intermittent note in one ear, the pitch of which would remain constant. Should the pitch become lower you would know that you had veered closer to the wall. If the pitch becomes higher you would have veered away from the wall. When you reach a point some 3 metres (9 ft.) from the buttress the device will stop displaying the side information and indicate the straight ahead object. But now the display is in both ears and at four times the repetition rate, thus grabbing your attention and warning you that evasive action must be taken.

Its artificial intelligence

The computer program which runs the Sonic Pathfinder includes, within the software, a number of sophistries which give the device a fairly high level of 'artificial intelligence'. The massive amount of information collected by the receivers is digitized and analyzed by a hierarchy of decision making algorithms which purport to select only that information relevant to the user's immediate needs. This selected information alone is displayed, thus, avoiding the informational overload commonly experienced by the users of some earlier devices.
The basic and simplest algorithm to understand is the one that gives priority to central objects. For example an object some 2 metres (6ft) away but in the direct line of travel would be displayed in preference to a shore line only one metre (3ft) to the side.
An early evaluation of the device showed that this algorithm alone did not reduce the displayed information sufficiently and clients were still finding difficulty when trying to negotiate narrow gaps. Careful observations of subjects using the device led to the hypothesis that much of the unwanted information would be eliminated if the device were made to have an overall range which changed automatically depending upon user need. In particular the overall range should be dependent on walking speed. The software algorithms to achieve this are complex. However, the new Sonic Pathfinder, instead of having a range of 2.5 metres (8 ft), has a range of 2 seconds! This is an unusual and odd concept. However, given the information processing demands inherent in independent blind travel and the moment to moment problem-solving nature of blind travel, it is highly desirable to have a device which is limited to providing information solely about those objects which would be encountered during the next 2 seconds of travel. What is more, the user is completely unaware of the massive amount of computation which is being perform on his or her behalf; the user hears a very simple display but one which is relevant to current needs.

The provision of training

The assumption that a simple-to-use device would require little training led to problems. In a series of experiments conducted in the Sonic Pathfinder; side close up United Kingdom devices were sent to experienced Mobility Instructors. The devices were accompanied by a detailed description but little advice was given about training. It was assumed that the Instructors would integrate the device into their own training strategy. This early attempt at distance learning failed; not one Instrutor felt sufficiently confident to train a vision impaired client. Only when a formal training course was developed was the situation rectified.
Considerable effort has been put into developing a training course suitable for teaching Mobility Instructors how to train clients in the use of the Sonic Pathfinder. Ideally Instructors attend a three day training course. To benefit from the course Instructors should be physically fit, experienced in the teaching of primary mobility devices and capable of independent travel under blind-fold. The course includes both theoretical and practical sessions. The Instructor Training Course manual is in the public domain. It is written in English and is 46 pages long in Word for Windows. To download a Zipped up version (approx. 560Kbytes) click on: Manual . Alternatively, a text version of the manual may be found on Tom Brennan's Page devoted to Electronic Travel Aids. The manual has been translated into German, French, Italian and Dutch.

In the light of the experience gained in developing and running the Instrucor Training Course a new revised Distance Learning Package has been developed for use my Mobility Instructors unable to attend a formal course.

Instructor training courses have been run in Australia, Indonesia, Germany, the UK and the USA. The device has been evaluated and approved by the US Department of Veterans' Affairs. There are now distributors and Accredited Sonic Pathfinder Instructors in Australia , UK, Europe , Indonesia and the USA .
Finally, to see some comments from a user; see what Graeme has to say.

Published Papers relating to the Sonic Pathfinder.

Heyes A.D. (1979)
Auditory information and the mobile blind.
PhD. Thesis, University of Nottingham, England.
Heyes A.D. (1981)
The use of musical scales to represent distance to object in an electronic travel aid for the blind.
Perceptual and Motor Skills. Vol.51 pp. 1015-1020.
Heyes A.D. (1983)
Human Navigation by Sound.
Physics in Technology. Vol.14 No.2 pp.68-75.
Heyes A.D. (1983)
The Sonic Pathfinder - A new travel aid for the blind.
In Technology aids for the disabled. Ed. W.J.Perkins. Pub. Butterworth. pp. 165-171.
Heyes A.D. (1984)
Sonic Pathfinder: a programmable guidance aid for the blind.
Electronics and Wireless World Vol.90, No.1579. pp.26-29 & 62.
Dodds A.G., Clark-Carter D.D. & Howarth C.I. (1984)
The Sonic Pathfinder: An evaluation.
(Carried out at the Blind Mobility Research Unit, University of Nottingham, England).
Journal of Visual Impairment & Blindness, 78 (5), pp. 306-310.
Heyes A.D. (1985)
Microprocessor techniques of information processing in electronic travel aids for the blind.
In Electronic Spatial Sensing for the Blind. Ed Warren & Strelow, NATO ASI series. pp 161-169.
Clark-Carter D.D., Heyes A.D. & Howarth C.I. (1986)
The effect of non-visual preview upon the walking speed of visually impaired people.
Ergonomics Vol. 29, No. 12, pp. 1575-1581.
Heyes A.D., Durinck M. & Beaton T. (1988)
The Sonic Pathfinder: Developments and preliminary field trail results.
In Orientation and Mobility of the Visually Impaired. Ed. Neustadt-Noy, Merlin and Schiff. Helliger Publishing Co. Jerusalem.
Clark-Carter D.D. & Heyes A.D. (1988)
The stress experienced by visually impaired pedestrians.
In Orientation and Mobility of the Visually Impaired. Ed. Neustadt-Noy, Merlin and Schiff. Helliger Publishing Co. Jerusalem.
Heyes A.D. & Clarke G. (1991)
The role of training in the use of the Sonic Pathfinder.
Proceeding of the American Association for the Education and rehabilitation of the Blind and Visually Impaired, Southwest Regional Conference, Hawaii.
Heyes A.D. (1994)
Sensory alternatives and the vision impaired.
Proceedings of the 7th. International Mobility Conference, Melbourne. RGDAA, Melbourne. pp. 9 - 12.
McKinley J., Goldfarb E. & Goodrich G. (1994)
An Evaluation of the Sonic Pathfinder.
(Carried out at the Western Blind Rehabilitation Center, Department of Veterans Affairs, Palo Alto, California USA).
Proceedings of the 7th. International Mobility Conference, Melbourne. RGDAA, Melbourne. pp. 177 - 179.
Gallimore D. & Beaton T. (1994)
Sonic Pathfinder: Utilised by 'Geoff', an intellectually and Vision impaired Client.
(Carried out at the Guide Dog Association of New South Wales, Australia).
Proceedings of the 7th. International Mobility Conference, Melbourne. pp.180 - 182.