How Do Auditors Order Their Tasks, and How Does Task Ordering Affect Performance Under Time Pressure?

I examine how the subjectivity of task criteria influences auditors' ordering and performance of audit tasks under time pressure. Tasks with more objective criteria provide little flexibility in how well they can be completed (i.e., they are either performed correctly or incorrectly). On the other hand, tasks with more subjective criteria have a wider range of performance levels that satisfy the "letter" of the criteria, but not necessarily the "spirit." I predict and find that auditors tend to work on a task with more objective criteria before a task with more subjective criteria. As time pressure increases, auditors ordering their tasks this way reduce performance on the subjective task, but not the objective task. By decreasing performance on tasks with more subjective criteria, auditors can address all the criteria for both tasks if only in letter, rather than in spirit.

DeepSoCS: A Neural Scheduler for Heterogeneous System-on-Chip (SoC) Resource Scheduling

In this paper, we present a novel scheduling solution for a class of System-on-Chip (SoC) systems where heterogeneous chip resources (DSP, FPGA, GPU, etc.) must be efficiently scheduled for continuously arriving hierarchical jobs with their tasks represented by a directed acyclic graph. Traditionally, heuristic algorithms have been widely used for many resource scheduling domains, and Heterogeneous Earliest Finish Time (HEFT) has been a dominating state-of-the-art technique across a broad range of heterogeneous resource scheduling domains over many years. Despite their long-standing popularity, HEFT-like algorithms are known to be vulnerable to a small amount of noise added to the environment. Our Deep Reinforcement Learning (DRL)-based SoC Scheduler (DeepSoCS), capable of learning the “best” task ordering under dynamic environment changes, overcomes the brittleness of rule-based schedulers such as HEFT with significantly higher performance across different types of jobs. We describe a DeepSoCS design process using a real-time heterogeneous SoC scheduling emulator, discuss major challenges, and present two novel neural network design features that lead to outperforming HEFT: (i) hierarchical job- and task-graph embedding; and (ii) efficient use of real-time task information in the state space. Furthermore, we introduce effective techniques to address two fundamental challenges present in our environment: delayed consequences and joint actions. Through an extensive simulation study, we show that our DeepSoCS exhibits the significantly higher performance of job execution time than that of HEFT with a higher level of robustness under realistic noise conditions. We conclude with a discussion of the potential improvements for our DeepSoCS neural scheduler.

Integrated real-time task and motion planning for multiple robots under path and communication uncertainties

SUMMARYWe consider a problem where robots are given a set of task locations to visit with coarsely known distances. The robots must find the task ordering that reduces the overall distance to visit the tasks. We propose an abstraction that models the uncertainty in the paths, and a Markov Decision Process-based algorithm that selects paths that reduces the expected distance to visit the tasks. We also describe a distributed coordination algorithm to resolve path conflicts. We have shown that our task selection is optimal, our coordination is deadlock-free, and have experimentally verified our approach in hardware and simulation.