Optimizing Locomotor Task Sets in Biological Joint Moment Estimation for Hip Exoskeleton Applications

Jimin An,Changseob Song,Eni Halilaj,Inseung Kang

2024-12-11

Abstract:Accurate estimation of a user's biological joint moment from wearable sensor data is vital for improving exoskeleton control during real-world locomotor tasks. However, most state-of-the-art methods rely on deep learning techniques that necessitate extensive in-lab data collection, posing challenges in acquiring sufficient data to develop robust models. To address this challenge, we introduce a locomotor task set optimization strategy designed to identify a minimal, yet representative, set of tasks that preserves model performance while significantly reducing the data collection burden. In this optimization, we performed a cluster analysis on the dimensionally reduced biomechanical features of various cyclic and non-cyclic tasks. We identified the minimal viable clusters (i.e., tasks) to train a neural network for estimating hip joint moments and evaluated its performance. Our cross-validation analysis across subjects showed that the optimized task set-based model achieved a root mean squared error of 0.30$\pm$0.05 Nm/kg. This performance was significantly better than using only cyclic tasks (p<0.05) and was comparable to using the full set of tasks. Our results demonstrate the ability to maintain model accuracy while significantly reducing the cost associated with data collection and model training. This highlights the potential for future exoskeleton designers to leverage this strategy to minimize the data requirements for deep learning-based models in wearable robot control.

Robotics,Machine Learning

What problem does this paper attempt to address?

### What problem does this paper attempt to solve?

This paper aims to solve the data collection challenges faced when estimating biological joint torques using deep - learning techniques. Specifically, the current state - of - the - art methods rely on deep - learning techniques, and these techniques require a large amount of laboratory data to train the model, which makes it very challenging to obtain sufficient data, especially when developing robust control strategies for lower - limb exoskeletons (particularly hip exoskeletons).

To address this challenge, the authors propose a **task - set optimization strategy** that aims to identify a minimal but representative task set to significantly reduce the data - collection burden while maintaining model performance. Through this method, researchers hope to significantly reduce the costs associated with data collection and model training without sacrificing model accuracy.

### Summary of main problems:

1. **Data - intensive problem**: Existing deep - learning methods require large and comprehensive datasets to represent various activity patterns and obtain real - joint torques through motion capture. This requirement leads to high costs and resource consumption.

2. **Data - collection bottleneck**: For certain specific populations (such as patient groups), data collection becomes more difficult due to physical limitations, time limitations, and logistical barriers.

3. **Lack of systematicity in task selection**: Although previous studies have verified that the optimal model performance can be achieved by reducing the task set, it has not been clear which tasks are the most relevant and useful, nor has it been explained why certain tasks are more conducive to developing user - independent joint - torque estimators.

### Solutions:

- **Task - set optimization**: Identify the most representative task set from multiple periodic and aperiodic tasks through clustering analysis and dimensionality - reduction techniques (such as PCA).

- **Neural - network training**: Use the optimized task set to train the neural network and evaluate its performance in estimating hip - joint torques.

### Results:

- The optimized - task - set model performs excellently in terms of root - mean - square error (RMSE) and coefficient of determination (\(R^{2}\)), comparable to the model using the complete task set, and significantly better than the model using only periodic tasks.

- This result indicates that by selecting a more representative task set, the accuracy of the model can be maintained while reducing the data - collection burden.

### Significance:

This study provides new ideas for future exoskeleton design, especially when developing user - independent hip - joint - torque estimators, which can effectively reduce the requirements for data collection and training, thereby promoting the application of exoskeletons in daily activities.