Preference-based assistance optimization for lifting and lowering with a soft back exosuit
Data files
Mar 28, 2025 version files 73.91 KB
-
LIG-JND-Dataset.zip
2.35 KB
-
LOG-JND-Dataset.zip
2.40 KB
-
Preference-Dataset.zip
67.32 KB
-
README.md
1.84 KB
Abstract
Wearable robotic devices have become increasingly prevalent in both occupational and rehabilitative settings, yet their widespread adoption remains inhibited by usability barriers related to comfort, restriction, and noticeable functional benefits. Acknowledging the importance of user perception in this context, this study explores preference-based controller optimization for a back exosuit that assists lifting. Considering the high mental and metabolic effort discrete motor tasks impose, we used a forced-choice Bayesian Optimization approach that promotes sampling efficiency by leveraging domain knowledge about just noticeable differences between assistance settings. Optimizing over two control parameters, preferred settings were consistent within and uniquely different between participants. We discovered that overall, participants preferred asymmetric parameter configurations with more lifting than lowering assistance, and that preferences were sensitive to user anthropometrics. These findings highlight the potential of perceptually guided assistance optimization for wearable robotic devices, marking a step towards more pervasive adoption of these systems in the real-world.
https://doi.org/10.5061/dryad.2z34tmpx5
Description of the data and file structure
This dataset comprises 3 zip files. One file each for the experiment to estimate a lowering gain/lifting gain (LOG/LIG) just noticeable difference (JND), respectively and one for the data of the main preference-based controller optimization experiment.
The JND experiments contain a csv file with the data for all subjects. The csv file contains 8 columns (all parameters are unitless), namely:
- Subject ID
- F1: LOG or LIG Option 1
- F2: LOG or LIG Option 2
- FB: User Feedback. 0/1 = Option 1/2 perceived as larger
- Block: Experiment Block
- LOG/LIG: The other control parameter which was held constant within a block
- gd: Gain difference, i.e., the difference between F1 and F2
- correct: Did subject correctly perceive the larger option as larger
The preference-based optimization experiment comprises separate csv files for each subject and experiment block. In addition, for each subject and block there exists a file ending in “Options.csv”, which contains the LOG/LIG pairs for each iteration and a file ending in “Feedback.csv” which contains the user feedback. More specifically, the Feedback.csv file has 12 rows (one for each iteration) and two columns. The numbers in the first column indicate which row in Options.csv was preferred and the numbers in the second column indicate the row index of the non-preferred option in that iteration.
Files and variables
File: LIG-JND-Dataset.zip
Description: Lifting gain JND dataset
File: LOG-JND-Dataset.zip
Description: Lowering gain JND dataset
File: Preference-Dataset.zip
Description: Preference dataset