Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Optimizing Task Completion Time Updates Using POMDPs

Duncan Eddy, Esen Yel, Emma Passmore, Niles Egan, Grayson Armour, Dylan M. Asmar, Mykel J. Kochenderfer

Abstract

Managing announced task completion times is a fundamental control problem in project management. While extensive research exists on estimating task durations and task scheduling, the problem of when and how to update completion times communicated to stakeholders remains understudied. Organizations must balance announcement accuracy against the costs of frequent timeline updates, which can erode stakeholder trust and trigger costly replanning. Despite the prevalence of this problem, current approaches rely on static predictions or ad-hoc policies that fail to account for the sequential nature of announcement management. In this paper, we formulate the task announcement problem as a Partially Observable Markov Decision Process (POMDP) where the control policy must decide when to update announced completion times based on noisy observations of true task completion. Since most state variables (current time and previous announcements) are fully observable, we leverage the Mixed Observability MDP (MOMDP) framework to enable more efficient policy optimization. Our reward structure captures the dual costs of announcement errors and update frequency, enabling synthesis of optimal announcement control policies. Using off-the-shelf solvers, we generate policies that act as feedback controllers, adaptively managing announcements based on belief state evolution. Simulation results demonstrate significant improvements in both accuracy and announcement stability compared to baseline strategies, achieving up to 75\% reduction in unnecessary updates while maintaining or improving prediction accuracy.

Optimizing Task Completion Time Updates Using POMDPs

Abstract

Managing announced task completion times is a fundamental control problem in project management. While extensive research exists on estimating task durations and task scheduling, the problem of when and how to update completion times communicated to stakeholders remains understudied. Organizations must balance announcement accuracy against the costs of frequent timeline updates, which can erode stakeholder trust and trigger costly replanning. Despite the prevalence of this problem, current approaches rely on static predictions or ad-hoc policies that fail to account for the sequential nature of announcement management. In this paper, we formulate the task announcement problem as a Partially Observable Markov Decision Process (POMDP) where the control policy must decide when to update announced completion times based on noisy observations of true task completion. Since most state variables (current time and previous announcements) are fully observable, we leverage the Mixed Observability MDP (MOMDP) framework to enable more efficient policy optimization. Our reward structure captures the dual costs of announcement errors and update frequency, enabling synthesis of optimal announcement control policies. Using off-the-shelf solvers, we generate policies that act as feedback controllers, adaptively managing announcements based on belief state evolution. Simulation results demonstrate significant improvements in both accuracy and announcement stability compared to baseline strategies, achieving up to 75\% reduction in unnecessary updates while maintaining or improving prediction accuracy.
Paper Structure (14 sections, 7 equations, 6 figures)

This paper contains 14 sections, 7 equations, 6 figures.

Table of Contents

  1. Introduction
  2. Related Work
  3. Background
  4. Partially Observable Markov Decision Processes
  5. Mixed Observability Markov Decision Processes
  6. Planning
  7. Announcement Control as a POMDP
  8. Results
  9. Solvers
  10. Baselines
  11. Simulation Results
  12. Reward Parameter Tuning: Pareto Analysis of $\lambda_e$ and $\lambda_c$
  13. Belief Evolution Analysis
  14. Conclusion and Future Work

Figures (6)

  • Figure 1: Comparison of reward values with different policies.
  • Figure 2: Comparison of average number of announcement changes with different policies.
  • Figure 3: Comparison of average increase in project completion time due to controller announcement time changes
  • Figure 4: Comparison of average error over the planning horizon with different policies. The value is computed as the average over all simulations of the sum total error across all planning steps.
  • Figure 5: Pareto frontier analysis showing the trade-off between average prediction error and number of announcement changes for qmdp policies. Each point represents a different $(\lambda_e, \lambda_c)$ parameter combination. The gold line indicates the Pareto-optimal configurations, while baseline policies are shown for reference. The black arrow provides the direction of global improvement.
  • ...and 1 more figures