Search

This paper presents a new multimodal control interface for people living with upper-body disabilities based on a wearable wireless sensor network. The proposed body-machine interface is modular and can be easily adapted to the residual functional capacities (RFCs) of different users. A custom data fusion algorithm has been developed for emulating a joystick control using head motion measured with a lightweight wireless inertial sensor enclosed in a headset. The wearable network can include up to six modular sensor nodes which can be used simultaneously to read different RFCs including gesture and muscular activity, and translate them into commands. Sensor data fusion is performed inside the sensor nodes in order to free the wireless link and the base station, and decrease power consumption. Requirements of such an interface are established for people using powered-wheelchairs, and a proof of concept system is implemented and used to control an assistive robotic arm. It is shown that the performance of the system compares well to conventional control systems like the joystick controller, while being potentially more suitable for the severely disabled.

This paper presents a new multimodal control interface for people living with upper-body disabilities based on a wearable wireless sensor network. The proposed body-machine interface is modular and can be easily adapted to the residual functional capacities (RFCs) of different users. A custom data fusion algorithm has been developed for emulating a joystick control using head motion measured with a lightweight wireless inertial sensor enclosed in a headset. The wearable network can include up to six modular sensor nodes which can be used simultaneously to read different RFCs including gesture and muscular activity, and translate them into commands. Sensor data fusion is performed inside the sensor nodes in order to free the wireless link and the base station, and decrease power consumption. Requirements of such an interface are established for people using powered-wheelchairs, and a proof of concept system is implemented and used to control an assistive robotic arm. It is shown that the performance of the system compares well to conventional control systems like the joystick controller, while being potentially more suitable for the severely disabled.

During overground wheelchair propulsion, upper body movement causes intra-cycle velocity variations that are neglected by current wheelchair simulators. This could affect the external validity of wheelchair propulsion on simulators. In this work, we investigated ways to incorporate these dynamics into the dynamic model (DM) reproduced by wheelchair simulators. We aimed to maximize the DM accuracy and minimize the number of required inputs. First, two DMs were presented: Model RL represented propulsion on a typical roller-based wheelchair simulator and model UB (upper body) represented overground propulsion, modelling the upper body as 5 rigid bodies. Then, three new DMs were presented: Model TR (trunk), model UA (upper arm) and model FA (forearm); these models simplified model UB by estimating the upper body kinematics based on the acceleration of only one segment. For all DMs, wheelchair velocity prediction was tested overground at a self-selected velocity among 19 experienced manual wheelchair users with a spinal cord injury. Upper body kinematics was reconstructed based on personalized kinematic patterns recorded on a wheelchair simulator. Models UB and UA were the most accurate: they reduced the root-mean-square intra-cycle velocity prediction error from 0.044 m/s (RL) to 0.026 m/s (UB) and 0.024 m/s (UA), and reduced the velocity peak time prediction error from -27.7 % (RL) to 1.7 % (UB) and -7.3 % (UA). Implementing model UA instead of model RL on a wheelchair simulator may improve the external validity of wheelchair propulsion on a simulator.

During overground wheelchair propulsion, upper body movement causes intra-cycle velocity variations that are neglected by current wheelchair simulators. This could affect the external validity of wheelchair propulsion on simulators. In this work, we investigated ways to incorporate these dynamics into the dynamic model (DM) reproduced by wheelchair simulators. We aimed to maximize the DM accuracy and minimize the number of required inputs. First, two DMs were presented: Model RL represented propulsion on a typical roller-based wheelchair simulator and model UB (upper body) represented overground propulsion, modelling the upper body as 5 rigid bodies. Then, three new DMs were presented: Model TR (trunk), model UA (upper arm) and model FA (forearm); these models simplified model UB by estimating the upper body kinematics based on the acceleration of only one segment. For all DMs, wheelchair velocity prediction was tested overground at a self-selected velocity among 19 experienced manual wheelchair users with a spinal cord injury. Upper body kinematics was reconstructed based on personalized kinematic patterns recorded on a wheelchair simulator. Models UB and UA were the most accurate: they reduced the root-mean-square intra-cycle velocity prediction error from 0.044 m/s (RL) to 0.026 m/s (UB) and 0.024 m/s (UA), and reduced the velocity peak time prediction error from -27.7 % (RL) to 1.7 % (UB) and -7.3 % (UA). Implementing model UA instead of model RL on a wheelchair simulator may improve the external validity of wheelchair propulsion on a simulator.