https://www.mdu.se/

Human motion analysis provides several important applications. Examples are fall risk assessment, sports biomechanics, physical activity monitoring and rehabilitation. This work in progress paper proposes an intelligent wireless body area network system for motion and gait symmetry analysis. A Bluetooth network with accelerometers, gyroscopes and in-shoe force sensing resistors gathers data and sends it to a web server after intelligent pre-processing and filtering. The system is flexible and adaptable for different use cases including combinations of gait analysis, gait symmetry and pressure measurements between foot and shoe.

Heavy working conditions, as well as sedentary behaviour, are risk factors for health. There is a lack of wearable measurement systems for monitoring carried loads while walking. Pedobarography, the study of force fields acting between the plantar surface of the foot and a supporting surface, is supposed to be useable for estimating carried loads. Purpose. The aim of this paper is to present a novel method for selecting appropriate measurement samples for weight estimation of carried load during walk and a wearable system, based on pedobarography, consisting of commercial off the shelf components. The main idea is to choose samples when half of the total weight is on the forward sensors and the other half is on the heel sensor “equipoise” in one foot while the other foot not touches the ground. Methods. The system consists of insoles with force sensing resistors, data acquisition with IOIO-OTG and analysis in Excel. Each subject was weighed on an electronic floor scale. Three walks were performed on level ground. The first walk without any added load and then with two increases of carried load. Equipoise was defined as having half the weight distributed on the heel and the other half over the metatarsal pad. An equipoise value of 0.5 represents equilibrium regarding the weight distribution on one foot, with the other foot in the air. Samples were chosen in the equipoise region of 0.5±0.1 and then the average of the samples collected during one minute estimated the total weight. Results. The system can detect increases in carried loads but has a tendency to overestimate them. The estimated value was always larger with increased weight but the system was not always linear. The average overestimation error was 16.7 kg. Discussion. This study shows that this type of wearable system is usable for estimating carried load during walk after calibration of the system to the body weight force distribution on the sensors. There is still need for future development to obtain real-time analysis and direct feedback. A smaller and lighter measurement system is also desirable. Conclusion. This study shows that the novel method, equipoise, is usable for selecting appropriate measurement samples for weight estimation of carried load during walk. This study also shows that the wearable system, consisting of commercial off the shelf components, can be used for these measurements. However, there is a tendency to overestimate the loads.

The aim of this pilot study was to investigate the possibility to find a correlation between the output from a portable pedobarography system and the walking intensity expressed as walking speed. The system uses shoe insoles with force sensing resistors and wireless transmission of the data via Bluetooth. The force-time integral, at the toe-off phase of the step, for the force sensors in the forward part of the right foot was used to measure impulse data for 10 subjects performing walks in three different walking speeds. This data was then corrected by multiplication with the step frequency. This pilot study indicates that the portable pedobarography system output shows a linear relationship with the walking intensity expressed as walking speed on an individual level.

Objectives

Walking speed is an important component of movement and is a predictor of health in the elderly. Pedobarography, the study of forces acting between the plantar surface of the foot and a supporting surface, is an approach to estimating walking speed even when no global positioning system signal is available. The developed portable system, Identifying Velocity and Load (IngVaL), is a cost effective alternative to commercially available pedobarography systems because it only uses three force sensing resistors. In this study, the IngVaL system was evaluated. The three variables investigated in this study were the sensor durability, the proportion of analyzable steps, and the linearity between the system output and the walking speed.

Methods

Data was collected from 40 participants, each of whom performed five walks at five different self-paced walking speeds. The linearity between the walking speed and step frequency measured with R2 values was compared for the walking speed obtained ‘A’ only using amplitude data from the force sensors, ‘B’ that obtained only using the step frequency, and ‘C’ that obtained by combining amplitude data for each of the 40 test participants.

Results

Improvement of the wireless data transmission increased the percentage of analyzable steps from 83.1% measured with a prototype to 96.6% for IngVaL. The linearity comparison showed that the methods A, B, and C were accurate for 2, 15, and 23 participants, respectively.

Conclusions

Increased sensor durability and a higher percentage of analyzed steps indicates that IngVaL is an improvement over the prototype system. The combined strategy of amplitude and step frequency was confirmed as the most accurate method.

Earlier work has shown that the IngVaL pedobarography system can estimate walking speed during indoor walking using only three forefoot sensors. The aim of this study was to examine if monitoring of walking speed using data from these three forefoot sensors is affected of the weight a person carries, if the person performs a walk in a set speed on the treadmill. Shoe insoles with force sensing resistors were connected to an electronic unit for signal conditioning and sampling and then the data was sent via Bluetooth to a tablet. Fifteen test persons walked five times each carrying five different weights on the treadmill at 1 m/s. The force-time integrals for the sum of the three forefoot sensors were calculated. This study shows that the force-time integrals for the three forefoot sensors shows a linear relationship with the carried weight as long as the person is not fatigued.

Personal health monitoring is advantageous in heavy work environments to reduce the risk of wear and tear and acute injuries. The study of forces between the plantar surface of the foot and a supporting structure, pedobarography, is a  promising candidate for monitoring carried weight during walk. The aim of this study was to evaluate the cost effective pedobarography measurement system, IngVaL. Two aspects are evaluated, namely, how well IngVaL can monitor carried weight during walk and if the novel implementation increased the durability. Fifteen test persons made five treadmill walks with a carried weight of 10, 20, 0, 15, and 5 kg. The equipoise analysis method was used. The Root Mean Square Error (RMSE) for estimation of the carried weight was 13.8 kg. A study with the earlier version of the measurement system had a RMSE of 23.3 kg. The earlier system, as well as commercial systems using this kind of sensors, have problems with sensor durability. The new sensor implementation, where the active sensor area boundary was no longer affected by mechanical stress, resulted in no broken sensors. This study shows an increased performance of carried weight estimation compared with earlier work, together with an improved sensor durability.