Hi. Disclosure up front: AI-assisted prose I reviewed before posting (https://github.com/URML-MARS/URML/blob/main/VIBE.md); human-only is fine if you prefer.
I maintain URML (https://urml.dev), a small Apache-2.0 language for declaring what a robot can do and checking an intended action against that declaration before it runs.
A planetary-analogue rover is a fitting case, because power and terrain bound what is admissible at any moment, and the loop that picks the next action benefits from a separate check that the action is within those bounds. The rover's autonomy decides where to go and what to do; URML answers whether the chosen action is admissible given a declared envelope: drive limits, manipulation reach, and a power-state or terrain constraint the mission declares. The check sits between action selection and the drivers, and leaves perception and the decision loop untouched.
The setting is part of the point: where a wrong action is costly and oversight is intermittent, a static pre-dispatch check earns its place.
Two questions:
- For your autonomy stack, is a declared drive-and-manipulation-and-power envelope check on a selected action useful before dispatch, or is feasibility already guaranteed inside action selection?
- Would a small worked example mapping a rover action onto a URML manifest, validated with no execution, be worth having?
Nothing here asks you to adopt or maintain anything. Thanks for the work, and good luck at the next rover competition.
Ido Yahalomi (greenvh@gmail.com)
Hi. Disclosure up front: AI-assisted prose I reviewed before posting (https://github.com/URML-MARS/URML/blob/main/VIBE.md); human-only is fine if you prefer.
I maintain URML (https://urml.dev), a small Apache-2.0 language for declaring what a robot can do and checking an intended action against that declaration before it runs.
A planetary-analogue rover is a fitting case, because power and terrain bound what is admissible at any moment, and the loop that picks the next action benefits from a separate check that the action is within those bounds. The rover's autonomy decides where to go and what to do; URML answers whether the chosen action is admissible given a declared envelope: drive limits, manipulation reach, and a power-state or terrain constraint the mission declares. The check sits between action selection and the drivers, and leaves perception and the decision loop untouched.
The setting is part of the point: where a wrong action is costly and oversight is intermittent, a static pre-dispatch check earns its place.
Two questions:
Nothing here asks you to adopt or maintain anything. Thanks for the work, and good luck at the next rover competition.
Ido Yahalomi (greenvh@gmail.com)