This month Aaron Chong will present on how to leverage the new MicroApp architecture of rmf-web to make custom dashboards for fleet management.
MicroApps are React components used to customize the rmf-web dashboard. This talk will go through what MicroApps are, how they can be built, and how to integrate them with an existing rmf-web dashboard. We will also go through some of the MicroApps that already come with the rmf-dashboard-framework, and how they interact with an Open-RMF deployment.
Do you think we can find a way to work together instead of duplicating work? I think we owe it to the ROS/robotics community to try and join efforts. Not only will we be able to move faster, we’ll also avoid confusion in terms of when to use which.
So far, the separation between Open-RMF and Transitive was that the former focused more on fleet coordination and infrastructure integration, while Transitive has always been focused on enabling users build their robot web dashboards for monitoring and control. I think this separation still makes sense and would love to see an integration of the two.
For reference, I believe rmf-web predates the public availability of Transitive, so its creation wasn’t a case of duplicating prior work. I think a key difference between Transitive and rmf-web is that rmf-web is first and foremost a reference implementation of how a dashboard can be integrated against Open-RMF. rmf-web is not particularly meant to be an out-of-the-box commercial product. Rather it’s meant to be a launching-off point for system integrators to see how to communicate with Open-RMF from a web context, and potentially reuse some of the relevant dashboard widgets.
The microapp update that Aaron presented was a rearchitecture of that reference implementation to improve its reusability and customizability. We still expect it to be a starting point in the journey of deploying a commercial fleet management dashboard. The rearchitecturing work was driven by a fleet deployment contract that was left over from OSRC. The customer for that contract was already familiar with rmf-web and specifically wanted to build off of it, so we focused on making that feasible for them rather than considering other commercial platforms.
In terms of consolidating efforts moving forward, there are several different approaches that come to mind.
One option could be to package up some of the existing rmf-web widgets in a way that they can be used in Transitive, but this will likely be redundant with many capabilities that already exist within Transitive.
Another option would be to introduce a bridge that converts between the rmf_api_msgs schemas and the schemas that Transitive uses.
Probably the best option (but also the most technically demanding on both ends) would be to converge on the schemas that are being used across both platforms.
Whatever way you would like to collaborate, a good starting point would be to share your ideas to the Open-RMF discussion board. We also have open project meetings and a community forum every other week (listed on the official OSRF events calendar) which you can join to ask questions participate in roadmap discussions.
