PERsonal NAvigation System for Visually disabled People
Background & Objectives
There are more than 13 million people in Europe who are visually impaired. They encounter serious difficulties in moving around and, until recently, could only rely on canes and dogs to travel more confidently in unfamiliar environments.
Fortunately some portable devices with their associated mobility services dedicated to disabled people have now appeared on the market. However, because they use only GPS receivers for localisation, they cannot provide the level of positioning accuracy and integrity needed by visually impaired people for assisting them efficiently in their mobility. The time to first fix appears to take too long and people are not confident enough in the reliability of the information they receive.
With EGNOS/Galileo capabilities, the PERNASVIP consortium is convinced that a considerable gain in precision could be achieved to reach the four-meter accuracy positioning asked for by the visually impaired associations. Service availability could also be largely improved in complex urban environments with EGNOS.
The use of EDAS data in the PERNASVIP project is the logical continuation of previous work realised by NAVOCAP on EGNOS data in 2003 under an ESA contract to experiment SISNET services in the context of public transport. On the one hand, the improvement of positioning will be reached by using correction data provided by EDAS, such as EGNOS augmentation messages adapted and sent via the GPRS link used to download A-GPS data. On the other hand, the improvement of positioning proposed by PERNASVIP in urban canyons will be reached using an innovative method for the treatment of GPS and MEMS signals. The algorithm developed by NAVOCAP uses raw data (pseudorange, integrated carrier PHASE, DOPPLER shift) delivered by a standard GPS and mixed with data coming from MEMS sensors (three gyrometers, three accelerometers and one three-axis magnetometer).
The development of a multimodal journey planner requires taking into account data coming from different sources (public transport databases, GIS, users’ databases).
The service will be tested by a minimum of 30 users in two pilot cities – Toulouse (FR) and Guipúzcoa (ES) – and validated both in terms of technical performance and users’ acceptance, with an objective of 75% satisfied pilot users.
This 18-month project will focus on improving an existing prototype of a portable guiding device with EGNOS/Galileo capabilities in three ways:
- increasing positioning accuracy and integrity;
- reducing the time to first fix;
- developing specific multimodal routing algorithms.
This device will be implemented in six verifiable objectives:
- define users’ requirements, with particular attention paid to the three ‘accuracy, reliability and service availability’ targets;
- improve the existing portable device with EGNOS/EDAS data treatment filters and algorithms in order to locate visually disabled pedestrians in urban environments to within 4 meters, 95% of the time, in less than 15 seconds time to first fix;
- develop multimodal itinerary computation algorithms which take into account walking constraints and optimisation parameters of importance for visually impaired pedestrians;
- design the appropriate IT infrastructure and interface in order to combine the application with multilayer data content such as public transportation databases, geographical information systems (GIS) or users’ community geo-referenced information;
- 30 users will test and validate the service in two pilot cities; the aim is to achieve a satisfaction rate of 75%;
- disseminate the project results and confirm the conditions for a successful business model.
Work performed & results
The results expected from the project are:
- defining the users’ requirements, which will have been addressed by the PERNASVIP prototype, while paying particular attention to the optimisation of the three ‘accuracy, reliability and service availability’ targets;
- improving the existing portable device with EGNOS/EDAS data treatment filters and algorithms so as to locate visually disabled pedestrians in urban environments with a time to first fix of less than 15 seconds and a 4-meter accuracy, 95% of the time;
- developing multimodal itinerary computation algorithms which take into account walking constraints and optimisation parameters of importance for visually impaired pedestrians (zebra crossings, not being able to turn around, large pavements, etc.);
- designing the appropriate IT infrastructure and interface in order to efficiently combine the application with multilayer data content such as public transportation databases, GIS or users’ community geo-referenced information enrichments.