Human Workload Prediction: Lag Horizon Selection

2025-05-21

AI Summary

Existing human–machine collaboration systems lack systematic investigation into optimal lag selection for human workload prediction, leading to suboptimal temporal responsiveness. Method: This paper proposes a lag-duration adaptive modeling framework that integrates wearable physiological signals (HRV, EDA) with task-context features via time-series analysis, constructing both univariate and multivariate predictive models. Contribution/Results: We first reveal a significant sensitivity difference in optimal lag duration between model types: ~240 s for univariate versus only ~120 s for multivariate models—enabling an adaptive lag-selection principle. In real-world human–machine collaborative tasks, our method achieves >85% accuracy in forecasting workload trends 30 seconds ahead, substantially improving prediction lead time and timeliness. It thus overcomes the traditional paradigm of reactive, instantaneous-feedback–driven control and establishes a reliable temporal window for proactive, dynamic human–machine strategy adaptation.

Technology Category

• Humans and AI: Planning and Decision Support for Human-Machine Teams
• Cognitive Modeling & Cognitive Systems: Simulating Human Behavior
• Machine Learning: Time-Series/Data Streams

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: Studies of user behavior, including longitudinal effects of personalized systems
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Human-robot teams must be aware of human workload when operating in uncertain, dynamic environments. Prior work employed physiological response metrics from wearable sensors to estimate the current human workload; however, these estimates only enable robots to respond to under- or overload conditions reactively. Current human workload prediction approaches are limited to short prediction horizons and fail to investigate variable lag horizons' impact on predictions. This letter investigates the impact of lag horizons on both univariate and multivariate time series forecasting models for human workload prediction. A key finding is that univariate predictions required longer lag horizons of 240 seconds (s), whereas multivariate workload predictions sufficed with shorter lag horizons with diminishing returns around 120s.

Problem

• Investigates lag horizons' impact on human workload prediction
• Compares univariate and multivariate time series forecasting models
• Determines optimal lag horizons for accurate workload prediction

Innovation

• Investigates lag horizons in workload prediction
• Compares univariate and multivariate forecasting models
• Identifies optimal lag horizons for different models