The project SYSIASS decomposes into three main activities:
1 / Prototype for intelligent wheelchair and autonomous navigation
This activity deals with the problem of the prototyping of autonomous and intelligent wheelchairs, which must have the capability of transporting people inside and outside buildings. The intelligent wheelchair must be able to localize itself in an a priori unknown environment, to detect the obstacles and to move safely from one point to another, to command wheelchair manually, semi-automatically or automatically depending on the state of the user. The movement orders can come either from the user or the embedded system. The wheelchair can also have specific functionalities depending on the user. Those functionalities must be defined according to the user’s wishes. For instance, some people involved in the project work jointly with hospitals and the organizations for disabled in order to identify the specific needs which those devices should fulfill.
The originality of this project is that the prototype can be used with most kind of wheelchair, so the designed device which brings intelligence to the wheelchair, is independent of its manufacturers.
This activity consists of the following major stages:
- Definition of the specificities and functionalities of the wheelchair prototypes.
- Instrumentation (sensors, actuators, computer interface) and data processing algorithms in order to achieve localisation and obstacle identification.
- System modelisation and off-line and on-line parameters identification.
- Design of path planning and robust trajectory tracking algorithms, under optimality criteria (minimum loss of energy, minimum time for displacement…) and under constraints (static and dynamic obstacles avoidance, keeping the communication links, limited information…).
- Elaboration of the navigation strategy in order to avoid the collision between the intelligent wheelchair and obstacles and to guarantee a safe functioning.
- Numerical implementation, tests and validation of the algorithms on the test bench.
To carry out this activity, each partner will bring its respective abilities. The complementarity of our expertise enables us to treat all the problems about autonomous navigation. Moreover, in this activity we will also decide the sensors required for the instrumentation of the prototype depending on the results of the survey to be conducted among the users with the occupational therapist involved in the project.
This activity is upstream based research, which leads to the involvement of several doctoral students (two from ISEN, one from Ecole Centrale Lille, one from University of Kent, and one from University of Essex)
2 / Access to medical data
To ensure maximal effectiveness of the autonomous wheelchair technology being developed, it is advantageous that a secure and efficient means of wireless communication be available between the wheelchairs and other locations such as a central medical database. Such advantages would include potentially introducing a high level control for the system, restricting the use of such technology to given times and locations and allowing timely personal medical data to be made available using the wireless medium. However, two major issues arise. Firstly, an efficient means of encrypting any transmitted data is required to prevent the potential compromise of personal medical data and ensure the data is only available in authorized locations. Secondly, some of the signals generated by the communicators may cause issues ranging from reducing operational effectiveness of medical equipment through to directly hazardous consequences for patients.
We therefore seek to develop novel technology to allow the widespread efficient and secure use of wireless mobile communication systems to be employed in a hospital environment without adversely affecting medical equipment such as defibrillators, infusion pumps, ventilators and cardiac monitors.
Special panels for integration into walls will be investigated to test (i) if they are capable of shielding rooms and equipment from unwanted signals and (ii) whether this can be achieved in such a way as to ensure improved access to medical data and convenience for users without adversely affecting other areas of hospital operation.
The proposed technology uses panels whose status may be varied to either allow signals to pass through them or not as required. Using a series of such panels located throughout a hospital, the availability of wireless signals within various locations may be controlled, shielding certain locations whilst allowing communication in others.
The dynamic nature of such panels also means the availability of a signal at a given location can be changed at short notice and even control which locations are able to communicate with others; for instance, night staff may require a different communications scenario to the day shifts.
To address the issue of secure communications, we note that while data encryption techniques are now highly sophisticated and well established, encryption itself cannot necessarily protect against fraudulent data manipulation when the security of encryption keys cannot be absolutely guaranteed. We therefore propose to enhance the use of an authentication system developed at the Universities of Kent and Essex based on the direct generation of encryption keys based on properties of electronic devices (termed ICmetrics). ICmetrics possess the dual advantage of not requiring the storage of any encryption key (which is regenerated as required) and also allows the identity of an electronic device to be uniquely determined, hence guaranteeing its authenticity.
3 / Patient – Wheelchair Interface
This activity is to continue our efforts on developing the advanced assistive technology for the disabled and elderly people to gain necessary mobility.
Our previous investigations have revealed that various hands-free human-machine interfaces will play a key role in enabling the disabled and elderly people to gain necessary mobility and independent living. Individuals who have limited control of their appendages may use their voice, gesture, eyeball movement, muscle and brain wave signals to communicate with computers and to control the wheelchairs and other assistive devices. However, these bio-signals are very noisy and no sensor is able to gather these signals accurately. Therefore, to build such hands-free human-machine interfaces for intelligent wheelchairs and other robotic devices are a non-trivial task.
In particular, the performance enhancement and algorithm optimisation of our hands-free human-machine interfaces for electric-powered wheelchairs will be conducted in this activity. We will focus on the integration of different hands-free human-machine interfaces (voice, gesture, brain-wave, muscle signals, eyeball and lip movements) and autonomous navigation functions that are developed in Activity 1. The developed hands-free controlled wheelchairs will be robust and intelligent, and has user-friendly man-machine interfaces and the ability of avoiding collision and planning a path. It will be able to help the elderly and disabled people to gain necessary mobility and independence, and have better quality of life in the society.
This activity completes the entire project and will involve all partners. It is the integration of the final system and will federate the work done in collaboration with each partner.