The task of producing Mars Science Laboratory (MSL) personnel schedules is daunting. Through automation, the labor and number of mistakes made are reduced when compared to the manual process. This software not only allows for common constraints to be expressed, but also for these to be gracefully descoped in case a solution cannot be found, thus making this unique in the personnel scheduling arena.
The primary improvement is the ability to gracefully retract constraints to achieve a schedule. Also, this is directly integrated into Excel, allowing for easier modification and utilization of the schedule.
Leveraging previous work on scheduling personnel for space mission operations, ASPEN (Activity Scheduling and Planning Environment) was adapted to the domain of scheduling personnel for operations of MSL. Skills indicate for what function the person is qualified to be scheduled. Off days are specific days that a person cannot be scheduled, e.g., vacation. Day-of-the-week constraints document which days of the week a person can be scheduled.
Shift-weighting is used to weight more demanding days more heavily when counting shifts. Duty cycle constraints limit the total number of shifts someone can be scheduled over a given period. There are three basic types of periods: shifts, days, and weeks.
For all of these constraints, the ability exists to descope to less restrictive constraints in a prescribed order should one fail to find a suitable schedule. Thus, for any individual, a collection of similar constraints is possible, each labeled with a descope level. Hard constraints (without descope annotation) can be associated as well.
Scheduling begins with the lowest descope level, and then attempts to find a schedule using squeaky-wheel optimization (SWO). Should SWO fail, then the descope level is increased, until either only hard constraints remain or a schedule is discovered. If only some of the tasks have been successfully assigned, the schedule with the lowest descope level with the most assigned tasks is returned.
The overall architecture of the system is simple. The schedule and constraints are part of an Excel spreadsheet. A macro saves the appropriate worksheet out as a comma-separated value (CSV) file and invokes the scheduler. The scheduler performs the necessary scheduling and then saves its results out as another CSV file. The macro then loads the newly created CSV file and copies the appropriate values into the worksheet.
This work was done by Russell L. Knight, Andrew H. Mishkin, Alicia R. Allbaugh, and Sharon L. Laubach of Caltech for NASA’s Jet Propulsion Laboratory. This software is available for commercial licensing. Please contact Dan Broderick at
This Brief includes a Technical Support Package (TSP).
Automated Scheduling of Personnel to Staff Operations for the Mars Science Laboratory
(reference NPO49283) is currently available for download from the TSP library.
Don't have an account?
Overview
The document titled "Automated Scheduling of Personnel to Staff Operations for the Mars Science Laboratory" outlines the development and implementation of an automated scheduling system designed to optimize personnel management for NASA's Mars Science Laboratory (MSL) operations. Created by a team from the Jet Propulsion Laboratory (JPL), the system leverages previous advancements in scheduling technologies, particularly the ASPEN (Activity Scheduling and Planning Environment) framework, to address the unique challenges of staffing for space missions.
The scheduling process is governed by various constraints that must be met to ensure effective personnel allocation. These constraints include skills, off-days (specific days when personnel cannot be scheduled), day-of-the-week availability, shift-weighting (which assigns different weights to shifts based on their difficulty), and duty-cycle constraints that limit the total number of shifts an individual can work over a specified period. For instance, in the MSL context, Fridays are weighted as two shifts due to the increased demands of planning for weekend operations.
The document emphasizes the need for a flexible scheduling system that can gracefully retract soft constraints when necessary to achieve a viable schedule. This is a critical feature that distinguishes the JPL system from other commercial scheduling solutions, such as ShiftPlanning, Time Forge Scheduling, and NimbleSchedule, which lack this capability.
The scheduling architecture is straightforward, utilizing an Excel spreadsheet to manage schedules and constraints. A macro within the spreadsheet facilitates the export of data to a CSV file, which is then processed by the scheduling algorithm. The results are saved back into the spreadsheet for easy access and review. The scheduling process employs a method known as Squeaky Wheel Optimization (SWO), which iteratively adjusts the descope levels of constraints until a suitable schedule is found or only hard constraints remain.
Overall, the document highlights the innovative approach taken by JPL to automate personnel scheduling for the MSL, showcasing the integration of advanced scheduling techniques and the importance of adaptability in meeting operational needs. This work not only enhances the efficiency of mission operations but also contributes to the broader field of automated planning and scheduling in aerospace applications.
