Embargoed until 2026-05-31
Author
Date
2023Type
- Doctoral Thesis
ETH Bibliography
yes
Altmetrics
Abstract
Manual wheelchairs are among the most commonly used devices in human rehabilitation. Their conventional design framework has stagnated for decades despite a large and growing population relying on these mobility aids and well-known related complications that include pressure ulcers, shoulder pain, and injuries. A major drawback is the high propulsion inefficiency since only a fraction of the energy exerted by the user is translated into wheelchair movement. In recent years, many projects have therefore aimed at improving the mechanical efficiency and decreasing the load on the shoulders, for instance by improving propulsion techniques or the wheelchair-user-combination.
However, a major source of energy loss is largely overlooked in the literature despite it being a direct and inherent consequence of conventional wheelchair-user-interaction principles. Wheelchairs commonly have 2 large wheels with push-rims for propulsion at the rear, and small castors at the front. All wheels are fully independent whereby the rear wheels rotate around a fixed axis and the front wheels follow the movement of the chair thanks to rotating castor forks with a small trail distance. A symmetric, bilateral push results in straightforward movement, a differential in rotational velocity between the rear wheels brings about a turn. This differential steering offers great flexibility and direct proprioceptive feedback, but results in the need for users to brake on one side in order to change or adjust direction which disrupts fluent and efficient movement.
This convention has profound consequences on users’ mobility. We found that wheelchair users change direction about 900 times a day whereby turns are frequently characterized by small turning radii in the range of 0.5 to 1m and small turning angles of less than 40°. Overall, wheelchair movement can be described as a concatenation of an initial orienting turn-on-the-spot and a series of straight lines and tight, correcting turns during movement. These polygonal movement patterns are most likely the result of users intuitively optimising energy efficiency as each turn, and especially larger bends, is associated with continuous braking on the curve-inner side, thereby exhausting propulsion energy and resulting in the need for users to re-accelerate after the turn. A similar effect can be observed during straightforward movement, especially on tilted surfaces such as most pavements. Here, the location of the centre of mass of the user forward of the rear axis provides a lever arm for gravity to rotate the wheelchair downhill resulting in wheelchairs veering off pavements and users being forced to continuously brake on the uphill side to counter this effect.
With steering-by-leaning, we propose novel concepts in the design of manual wheelchairs that aim at removal of these adverse conditions. Such systems are actuated by a mobile backrest that acts as a steering wheel, directionally controls the front wheels, and guides the movement of the entire chair. Fundamentally, any steering mechanism could allow wheelchair users to change or adjust the direction of travel without braking, increase overall efficiency, and therefore decrease the strain on the shoulders. Using trunk movement as an input gesture to control steering, further combines propulsion efficiency with hand freedom while moving, stimulation of core blood flow and digestion as well as improved back health and trunk stability as promising side effects of more dynamic sitting.
Of course, trunk control and sitting stability are often limited among wheelchair users and this proposals’ applicability critically depends on users’ ability to accurately control the steering input. Steering-by-leaning systems are therefore either only applicable for people with full trunk functionality or critically need to support trunk movement dynamically, and consider individual functional abilities.
Among several possible designs for appropriate dynamic backrests, a simple 2D hinge joint to approximate lateral flexion in the lumbar spine appeared to be sufficient. Critically, after individually adjusting the backrest range of motion, the stiffness of springs to support users returning to an upright sitting posture, as well as the translation ratio of backrest to steering angles, we found that the system appeals to many and that the functional group of potential users even includes people with complete tetraplegia. Adapting the system’s behaviour to different individual requirements, possibly even accommodating for functional changes over time would seem simple with a modularised electronic steering-by-wire system as implemented during this project for research purposes. The substantial added bulk and weight of motors and the dependency of users on charged batteries, however, suggested that mechanical solutions will be preferred.
A “study prototype” of a steering-by-leaning wheelchair forms a central development result of this project. It was built by implementing findings from studying wheelchair movement in daily life, assessing the trunk functional spectrum among active wheelchair users as well as testing early, modularised prototypes with users. Besides a mechanical steering linkage based on flexible and readily available Bowden cables, it was equipped with various sensors to quantify effects of a steering-by-leaning system on wheelchair propulsion biomechanics. A set of measurement wheels were essential instruments to track participants’ positive and negative power input bilaterally. Here, a novel sensor layout based on 4 2D load cells that connect the push-rims to the wheel-rims allowed for a lightweight and compact design that minimises the inertial influence of the instruments on propulsion patterns.
To compare propulsion biomechanics of a steering-by-leaning against a conventional mode, experienced wheelchair users as well as able-bodied novices were asked to repeatedly complete a standardised agility test course that included straight level as well as tilted sections in addition to slalom and 180°-turning trajectories. The results suggest that the portion of negative propulsion power, that is, the role of braking in functional wheelchair movement, is, indeed, integral. The steering-by-leaning system, however, renders braking obsolete and therefore allowed users to save 35% (wheelchair users) to 50% (able-bodied) of energy despite travelling faster and further. Steering-by-leaning appears to drastically increase the efficiency of wheelchair propulsion, not by increasing the fraction of force that is translated into wheelchair movement but by allowing users to better utilise kinetic energy. Together with intuitive movement control and “fun factor”, the reduced physical demand of wheelchair propulsion promises to decrease the strain on users’ shoulders while encouraging healthy physical activity. The further development of steering-by-leaning systems is therefore clearly indicated towards contributing to wheelchair users’ health, mobility, and independence.
The research and development of this applied project therefore ends with a preliminarily final design of a designated wheelchair prototype with mechanical and fully integrated steering system is ready to be challenged in daily life environments. The use of cam-follower mechanisms allows concentrating the complexity of individually configuring the system’s behaviour into one simple component whereby the core mechanisms remain standard across different chairs. Future work should primarily aim at the translation of the results into tangible products but should critically focus on the development of reliable methods for individualisation. Further studies are needed to establish optimal steering and propulsion techniques and to longitudinally investigate effects of steering-by-leaning on users’ physical activity levels and health. Show more
Permanent link
https://doi.org/10.3929/ethz-b-000614492Publication status
publishedExternal links
Search print copy at ETH Library
Publisher
ETH ZurichSubject
Manual Wheelchair Design; Rehabilitation Engineering; Assistive Technology; Steering-by-LeaningOrganisational unit
02518 - Institut für Biomechanik / Institute for Biomechanics
More
Show all metadata
ETH Bibliography
yes
Altmetrics