Hi GTSAM maintainers,
URML (urml.dev) is a small, Apache-2.0 language for describing robot intent: English sentence -> typed primitive -> static validation against a capability manifest and a safety envelope -> dispatch. URML declares frames and named locations; it consumes a pose estimate to ground them, it does not estimate anything itself.
I want to be upfront that GTSAM is the lowest-direct-fit target in this round, and I am posting partly to get the boundary right rather than to claim a mapping. GTSAM is the factor-graph backend that a lot of SLAM and state-estimation systems optimize on; URML sits far above that. The honest relationship is indirect: the estimate a GTSAM-backed system produces is what ultimately grounds URML's frames.
So one real question, genuinely a question: is there any sensible point of contact between a thin intent/spec layer like URML and a factor-graph optimization backend, or is the right answer that URML should only ever talk to the system built on top of GTSAM (an estimator), never to GTSAM itself? A clear "stay one layer up" is a useful answer.
Full write-up: https://github.com/URML-MARS/URML/blob/main/docs/rfcs/0333-gtsam-outreach.md
One small thing: the GitHub API did not surface an SPDX license id at our verification time. Is GTSAM BSD-licensed?
Thanks for GTSAM; it is foundational to a remarkable amount of robotics.
Ido Yahalomi (URML, greenvh@gmail.com)
AI-assisted prose, maintainer-reviewed before posting (see VIBE.md). Human-only correspondence available on request.
Hi GTSAM maintainers,
URML (urml.dev) is a small, Apache-2.0 language for describing robot intent: English sentence -> typed primitive -> static validation against a capability manifest and a safety envelope -> dispatch. URML declares frames and named locations; it consumes a pose estimate to ground them, it does not estimate anything itself.
I want to be upfront that GTSAM is the lowest-direct-fit target in this round, and I am posting partly to get the boundary right rather than to claim a mapping. GTSAM is the factor-graph backend that a lot of SLAM and state-estimation systems optimize on; URML sits far above that. The honest relationship is indirect: the estimate a GTSAM-backed system produces is what ultimately grounds URML's frames.
So one real question, genuinely a question: is there any sensible point of contact between a thin intent/spec layer like URML and a factor-graph optimization backend, or is the right answer that URML should only ever talk to the system built on top of GTSAM (an estimator), never to GTSAM itself? A clear "stay one layer up" is a useful answer.
Full write-up: https://github.com/URML-MARS/URML/blob/main/docs/rfcs/0333-gtsam-outreach.md
One small thing: the GitHub API did not surface an SPDX license id at our verification time. Is GTSAM BSD-licensed?
Thanks for GTSAM; it is foundational to a remarkable amount of robotics.
Ido Yahalomi (URML, greenvh@gmail.com)
AI-assisted prose, maintainer-reviewed before posting (see VIBE.md). Human-only correspondence available on request.