we testing some route optimiztion problem.
in sometime we need route overlap blocking(edge by edge or ALL edge)
is there any option for route overlap blocking?
thx
@telos.l welcome, thanks for your question!
Can you help us understand your use case a little more? Are you asking if there is a way to prevent vehicles from traveling along the same path in a waypoint graph, or if there is a way to prevent more than one vehicle from visiting a location, or even that a location is only visited once by a single vehicle? Or something else?
Currently we don’t have such restrictions, cuOpt tries to optimize for the fewest number of vehicles as the primary default objective. However, the engineering team has thought about the possibility of spatio-temporal restrictions as an area of reseach. Understanding your use case would be helpful.
Thanks!
@user162039 thanks for your reply!
In our case is : there is a way to prevent more than one vehicle from visiting a location
(For example, some roads are so narrow that one vehicle cannot pass at the same time as another)
May I ask how far your team’s development of the temporal and spatial constraints is going? Or how is it going?
Thanks!
Hi there,
Having the possibility to introduce spatial constraints into the optimization problem also has my great interest.
@user162039 Has the engineering team been able to realize their thoughts regarding such constraints into cuOpt? I would be happy to perform testing and deliver feedback. In a later stage, I could also provide benchmarking.
Best regards,
Yannick
Hi @yannick.vandijk,
Thank you for the offer! There are no concrete features around this yet or a published roadmap, but the team is still thinking and learning about it and interested in different perspectives. If you have references to related white papers or other applications that do this, or descriptions of use cases, we’d love to see them.
One intermediate step the team has considered is to allow estimation and labeling of congested areas as a constraint on the problem so that those areas could be avoided.
Hello @user162039,
Thank you for your quick reply. It is good to hear that the development team is still interested in looking into this feature.
At the moment, I cannot provide any white papers or similar applications. It is, however, a topic for which we are currently developing an in-house solution. The benchmarking would be done on this basis.
Regarding the use case; the solver is used to optimize the cycle time of robot cells that contain multiple industrial manipulators. In order to ensure a collision-free solution, the solver must respect specific spatial constraints. The spatial constraints, defined for every edge in the graph, is represented as a bitset (00010101…) for which every bit represents a voxel within the robot cell.
Naturally, the data format of the spatial constraints can be adjusted, but from my point of view a bitset allows for the quickest computations.
Best regards,
Yannick