Locomotion Mode Transitions: Tackling System- and User-Specific Variability in Lower-Limb Exoskeletons

Andrea Dal Prete,Zeynep Özge Orhan,Anastasia Bolotnikova,Marta Gandolla,Auke Ijspeert,Mohamed Bouri

2024-11-19

Abstract:Accurate detection of locomotion transitions, such as walk to sit, walk to stair ascent, and descent, is crucial to effectively control robotic assistive devices, such as lower-limb exoskeletons, as each locomotion mode requires specific assistance. Variability in collected sensor data introduced by user- or system-specific characteristics makes it challenging to maintain high transition detection accuracy while avoiding latency using non-adaptive classification models. In this study, we identified key factors influencing transition detection performance, including variations in user behavior, and different mechanical designs of the exoskeletons. To boost the transition detection accuracy, we introduced two methods for adapting a finite-state machine classifier to system- and user-specific variability: a Statistics-Based approach and Bayesian Optimization. Our experimental results demonstrate that both methods remarkably improve transition detection accuracy across diverse users, achieving up to an 80% increase in certain scenarios compared to the non-personalized threshold method. These findings emphasize the importance of personalization in adaptive control systems, underscoring the potential for enhanced user experience and effectiveness in assistive devices. By incorporating subject- and system-specific data into the model training process, our approach offers a precise and reliable solution for detecting locomotion transitions, catering to individual user needs, and ultimately improving the performance of assistive devices.

Robotics

What problem does this paper attempt to address?

The problem that this paper attempts to solve is: **How to accurately detect the transition of motion patterns in lower - limb exoskeletons to adapt to different users and system characteristics, thereby improving the control effect of assistive devices and user experience**.

Specifically, the paper points out that in practical applications, due to individual differences among users (such as gait changes) and differences in exoskeleton designs, traditional non - adaptive classification models are difficult to maintain high - precision transition detection and are prone to delays. To overcome these problems, the authors propose two methods to adapt to system and user - specific changes: a statistics - based method and a Bayesian optimization method. These methods aim to improve the transition detection accuracy among different users by individually adjusting the thresholds of the finite - state - machine (FSM) classifier.

The following are the key issues mentioned in the paper:

1. **Accurate detection of motion - pattern transitions**:
- Accurately detecting the transitions from walking, going up and down stairs to sitting is crucial for effectively controlling robotic assistive devices (such as lower - limb exoskeletons).
- Each motion pattern requires specific assistance, so real - time identification of transitions is very important.

2. **Variability in user and system specificity**:
- Changes in user behavior (such as gait differences) and differences in exoskeleton mechanical designs will introduce variability in sensor data, making it difficult to maintain high - precision transition detection.
- Traditional methods perform poorly in the face of dynamic, real - world scenarios, especially when different terrains and diverse user behaviors are involved.

3. **Limitations of existing methods**:
- Most of the existing methods focus on the detection of specific stable - state activities and ignore the transition detection between different activities.
- These methods may fail in the face of more complex real - world scenarios, especially when there are large differences between different exoskeleton designs and user groups.

To solve the above problems, the authors propose the following solutions:

- **Statistics - based method**:
- Adjust the FSM threshold by analyzing the statistical differences between training data and specific - user data.
- Use the formula \(\text{TH}_{\text{new}}=\text{TH}_{\text{tr}}\cdot\frac{\text{mean}(\text{ICF}_{\text{new}})\pm\text{std}(\text{ICF}_{\text{new}})}{\text{mean}(\text{ICF}_{\text{tr}})\pm\text{std}(\text{ICF}_{\text{tr}})}\) to adjust the threshold, where \(\text{TH}_{\text{tr}}\) is the old threshold obtained from offline training, and \(\text{TH}_{\text{new}}\) is the threshold after statistical adjustment.

- **Bayesian optimization method**:
- Use Bayesian optimization techniques to optimize the FSM threshold according to individual - user data.
- By minimizing the experimental time, reduce the long - term physical consumption of users, which is especially important for people with limited mobility.

These two methods significantly improve the transition detection accuracy among different users, emphasize the importance of personalization in adaptive control systems, and ultimately improve the performance of assistive devices and user experience.