Retention of Manual Control Skills in Multi-Axis Tracking Tasks

Abstract

This paper presents the results of a training and retention experiment conducted to objectively and quantitatively evaluate the acquisition, decay, and retention of skill-based manual control behavior in a compensatory dual-axis roll and pitch attitude tracking task. In this study, thirty-eight fully task-naive participants were trained in a fixed-base setting in the Human-Machine Interaction Laboratory at Delft University of Technology and subsequently divided into three matched groups based on their training performance and control behavior. Performance of the first group was re-evaluated after a period of non-practice of six months, whereas the second group was retested at both three and six months after training, and skill retention of the third group was measured after two, four, and six months. The goal of the experiment was to model the decay curve of skill-based manual control behavior and to determine the re-acquisition rate of lost skills compared to their initial acquisition rate. To explicitly quantify changes in manual control skills, learning curve models were fitted to metrics of task performance and control activity. The results suggest that control skills decay following a negatively accelerating decay curve and that lost skills are re-acquired at a higher rate than their initial development rate.

Prediction Models for Individuals' Control Skill Development and Retention using XGBoost and SHAP

“…In this paper, data from a dual-axis compensatory tracking task experiment is used to develop a model that can predict individual skill decay over a period of inactivity. The experiment is described in [20], where its data is used for the objective evaluation of the retention of manual control skills using a 'cybernetic' pilot modeling method. To acquire operationally relevant results, a dual-axis pitch/roll tracking task was performed by a total of 43 task-naive participants, of whom 37 were able to provide a complete dataset.…”

“…The HO controls the 𝜙 and 𝜃 outputs of the aircraft dynamics 𝐻 𝑐 𝜙 and 𝐻 𝑐 𝜃 , respectively, using a control stick with roll and pitch gains 𝐾 𝑠 𝜙 and 𝐾 𝑠 𝜃 . Human manual control behaviour in such a dual-axis compensatory tracking task can be captured with two parallel error responses [20,21], as indicated in Fig. 2 by 𝐻 𝑝 𝑒 𝜙 (𝑠) and 𝐻 𝑝 𝑒 𝜃 (𝑠).…”

“…In [20], measured HO control behavior was quantified using "cybernetic" HO models as proposed in [21,22]. In compensatory tracking tasks, see also Fig.…”

“…( 1) and ( 2), 𝐾 𝑝 , 𝑇 𝐿 , 𝜏 𝑒 , 𝜔 𝑛𝑚 and 𝜁 𝑛𝑚 are the HO's control gain, lead time-constant, time delay, neuromuscular frequency, and neuromuscular damping ratio, respectively. In [20], these parameters were estimated separately for both the roll and pitch human operator responses shown in Fig. 2.…”

“…The participants in Groups 2 and 3 returned after 90 days or 60 and 120 days of inactivity, respectively, for additional intermediate retention tests. Finally, for potential correlation with the measured retention performance data [20] accumulated demographic data of all participants through a pre-experiment survey. However, no statistically significant correlation at the group level was identified, see the corresponding Appendix of [24] for details.…”