Hi!
We have done several projects that involve the use of omnidirectional AMRs, which means they can perform full vehicle rotations to continue traveling longitudinally or rotate the drive unit to switch from longitudinal to transverse mode, or vice versa.
On one hand, the fact that an AMR travels longitudinally or transversely affects the speed and acceleration characteristics, as well as the effective length of the task executer.
On the other hand, the complete rotation time and the rotation time of the drive unit are significantly different. When comparing scenarios using the maximum rotation time versus an average value, a significant difference is observed in the number of AMRs required to meet the system's needs.
Their navigation system is capable of deciding whether it is beneficial for an AMR to perform one type of rotation or another, based on the time it will take to travel in one way or another. As a simplification, in most models, we assume that there are specific paths where it will travel transversely, and we assign a higher speed to those paths (it would also be good to be able to assign different accelerations and decelerations depending on the type of path). However, even with this assumption, we cannot consider different rotation times or orientations as required.
Have you considered a system with these characteristics? Is it possible for you to develop a functionality for omnidirectional AMRs that attempts to cover these modeling needs, at least the ability to consider different rotation times and respect the orientation if we assign which paths will be traveled longitudinally and which ones transversely?
I am attaching a model where we have done some tests to try to work with these aspects during the travels of the task executers and the issues we have encountered.
23.1 Omnidirectional AMRs TEST.fsm
Thank you very much.
Best regards.
