BackgroundThe Mars 2020 rover is scheduled for launch in the summer of 2020 and arrival at Mars in February 2021. The current plan is to have an onboard scheduler help the rover independently adapt to execution variances on the surface of Mars, enabling the rover to operate more efficiently. Specifically, if activities finish earlier or later than expected, the onboard scheduler will allow the rover to extend or add activities to perform more science, or delay activities to be performed later.
ProblemEfficiently operating a rover on the surface of Mars is challenging. Two factors combine to make this job particularly difficult: 1) communication opportunities are limited, 2) certain aspects of rover performance are difficult to predict. With limited communications, the rover must be given instructions on what to do for one or more Martian days at a time. In addition, the duration of many rover activities can be hard to predict, which leads to unpredictable energy use. Traditionally, conservatism is used to keep the rover safe and healthy. This approach, however can lead to a measurable loss in rover productivity. To regain some of this productivity, the Mars 2020 mission is prototyping the use of onboard scheduling software. The primary objective of this software is to identify and utilize opportunities that arise when actual rover performance is more efficient than the original, conservative prediction.
ImpactAn analysis of Mars Science Laboratory operations [Gaines et al. 2016] indicates that significant onboard resources are made available during execution that cannot be used due to current rovers inability to change the uploaded schedule. If an onboard scheduler can leverage such resources, it can significantly improve rover productivity.
StatusAs of Summer 2019 the onboard software and supporting ground software are in development for operations in late 2021.
TeamDr. Daniel Gaines
Dr. Steve Chien