Automatic wheelchairs have evolved in terms of instrumentation and control, solving the mobility problems of people with physical disabilities. With this work it is intended to establish the background of the instrumentation and control methods of automatic wheelchairs and prototypes, as well as a classification in each category. To this end a search of specialised databases was carried out for articles published between 2012 and 2019. Out of these, 97 documents were selected based on the inclusion and exclusion criteria. The following categories were proposed for these articles: (a) wheelchair instrumentation and control methods, among which there are systems that implement micro-electromechanical sensors (MEMS), surface electromyography (sEMG), electrooculography (EOG), electroencephalography (EEG), and voice recognition systems; (b) wheelchair instrumentation, among which are found obstacle detection systems, artificial vision (image and video), as well as navigation systems (GPS and GSM). The results found in this review tend towards the use of EEG signals, head movements, voice commands, and algorithms to avoid obstacles. The most used techniques involve the use of a classic control and thresholding to move the wheelchair. In addition, the discussion was mainly based on the characteristics of the user and the types of control. To conclude, the articles exhibited the existing limitations and possible solutions in their designs, as well as informing the physically disabled community about the technological developments in this field.

The design and implementation of an electronic system that involves head movements to operate a prototype that can simulate future movements of a wheelchair was developed here. The controller design collects head-movements data through a MEMS sensor-based motion capture system. The research was divided into four stages: First, the instrumentation of the system using hardware and software; second, the mathematical modeling using the theory of dynamic systems; third, the automatic control of position, speed, and orientation with constant and variable speed; finally, system verification using both an electronic controller test protocol and user experience. The system involved a graphical interface for the user to interact with it by executing all the controllers in real time. Through the System Usability Scale (SUS), a score of 78 out of 100 points was obtained from the qualification of 10 users who validated the system, giving a connotation of “very good”. Users accepted the system with the recommendation to improve safety by using laser sensors instead of ultrasonic range modules to enhance obstacle detection.

In the last ten years, the design and implementation of Optical Fiber Sensors (OFS) in biomedical applications have been discussed, with a focus on different subareas, such as body parameter monitoring and control of assistive devices. A scoping review was performed including scientific literature (PubMed/Scopus, IEEE and Web of Science), patents (WIPO/Google Scholar), and commercial information. The main applications of OFS in the rehabilitation field for preventing future postural diseases and applying them in device controllers were discussed in this review. Physical characteristics of OFS, different uses, and applications of Polymer Optical Fiber pressure sensors are mentioned. The main postures used for posture monitoring analysis when the user is sitting are normal position, crooked back, high lumbar pressure, sitting on the edge of the chair, and crooked back, left position, and right position. Additionally, it is possible to use Machine Learning (ML) algorithms for posture classification, and device control such as Support Vector Machine, k-Nearest Neighbors, etc., obtaining accuracies above 97%. Moreover, the literature mentions wheelchair controllers and Graphical User Interfaces using pressure maps to provide feedback to the user. OFS have been used in several healthcare applications as well as postural and preventive applications. The literature showed an effort to implement and design accessible devices for people with disabilities and people with specific diseases. Alternatively, ML algorithms are widely used in this direction, leaving the door open for further studies that allow the application of real-time systems for posture monitoring and wheelchairs control.

Recently, robotic wheelchairs commanded by Human-Machine Interfaces (HMIs) have gained recognition for enhancing the quality of life of people with physical disabilities. In this sense, implementing sensors that allow for accurate and comfortable user intention recognition remains challenging. Additionally, posture monitoring for pressure ulcer prevention in wheelchair users is often overlooked. In this study, three HMIs are proposed to recognize the user’s intention using information linked to head, and neck movements, and visually evoked potentials to control an electric-powered wheelchair in four directions: forward, left, right, and back. These HMIs incorporate technologies, such as a IMU-based system, pressure sensors based on Polymeric Optical Fiber (POF), and steady-state visual evoked potential-based brain-computer interface (BCI) to generate control commands. The POF-based pressure sensors also allow for posture classification. The HMIs were evaluated functionally, and the user’s experience was considered from healthy subjects. The head-motion-based system obtained the highest Accuracy (ACC) rate (~0.99) and less workload. In contrast, the BCI reached the highest satisfaction and usability, whereas the neck-motion-based system achieved the lowest latency (~28 ms). The posture classification system achieved an acceptable ACC (~0.80), latency (~117 ms), and good perception. These results have great implications for the design of wheelchair systems to improve the independence of people with reduced mobility using information from multiple sources and for posture monitoring towards the prevention of pressure ulcers generation.

A proof of concept of a wheelchair posture monitoring based on Polymer Optical Fiber (POF) pressure array system using machine learning algorithms was made. Twelve lines of sensors were implemented in a wheelchair and ten postures were analyzed using classification algorithms. The classification system used a CAR filter and a SVM algorithm leading to an accuracy of 99.52%, and a cross validation of 98.83%. The best cross validation was 99.16% using a kNN algorithm.