RobotLocomotion / gcs-science-robotics

Motion Planning around Obstacles with Convex Optimization by Marcucci et al, 2023
BSD 3-Clause "New" or "Revised" License
181 stars 28 forks source link

Trouble with Regenerate IRIS Regions for bimanual_iiwa_example.ipynb #9

Open Larkin-AP opened 6 months ago

Larkin-AP commented 6 months ago

Hello, I am having problems running bimanual_iiwa_example.ipynb in docker.

When Regenerate IRIS Regions is unmodified, errors are displayed on the system as follows: Failure at geometry/optimization/iris.cc:759 in IrisInconfigqurationspace(): condition 'p.PointInset(sample, le-12)’ failed.

As shown in the figure 1

In fact, only this seed throws the error, and the entire program fails to exit the thread pool properly.

I saw that seed_points were generated from getConfigurationSeeds. The positions in getConfiguration define the joint angles of the robotic arm at some positions. Are these configuration positions the end points of the target trajectory?

Look forward to your reply and thank you for your help!

Larkin-AP commented 6 months ago

hi! I tried to slightly modify the faulty seed to avoid this problem for the time being. However, the same Regions, when running the program often times out, occasionally can run successfully, may I ask why? Thank you very much for your answer!

RussTedrake commented 6 months ago

My apologies! We've made substantial improvements to the entire pipeline. I will try to port this example over to the latest Drake implementations in the next few days.

hanlanqian commented 1 month ago

Hello, I am having problems running bimanual_iiwa_example.ipynb in docker.

When Regenerate IRIS Regions is unmodified, errors are displayed on the system as follows: Failure at geometry/optimization/iris.cc:759 in IrisInconfigqurationspace(): condition 'p.PointInset(sample, le-12)’ failed.

As shown in the figure 1

In fact, only this seed throws the error, and the entire program fails to exit the thread pool properly.

I saw that seed_points were generated from getConfigurationSeeds. The positions in getConfiguration define the joint angles of the robotic arm at some positions. Are these configuration positions the end points of the target trajectory?

Look forward to your reply and thank you for your help!

bro, I got the same problem, have u ever solved this problem?

Larkin-AP commented 1 month ago

Sorry, my friend. I haven't solved this problem yet, and I don't know if it's the cause of unstable solution.

At 2024-09-25 21:48:13, "Elaina" @.***> wrote:

Hello, I am having problems running bimanual_iiwa_example.ipynb in docker.

When Regenerate IRIS Regions is unmodified, errors are displayed on the system as follows: Failure at geometry/optimization/iris.cc:759 in IrisInconfigqurationspace(): condition 'p.PointInset(sample, le-12)’ failed.

As shown in the figure

In fact, only this seed throws the error, and the entire program fails to exit the thread pool properly.

I saw that seed_points were generated from getConfigurationSeeds. The positions in getConfiguration define the joint angles of the robotic arm at some positions. Are these configuration positions the end points of the target trajectory?

Look forward to your reply and thank you for your help!

bro, I got the same problem, have u ever solved this problem?

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

hanlanqian commented 1 month ago

Sorry, my friend. I haven't solved this problem yet, and I don't know if it's the cause of unstable solution. At 2024-09-25 21:48:13, "Elaina" @.> wrote: Hello, I am having problems running bimanual_iiwa_example.ipynb in docker. When Regenerate IRIS Regions is unmodified, errors are displayed on the system as follows: Failure at geometry/optimization/iris.cc:759 in IrisInconfigqurationspace(): condition 'p.PointInset(sample, le-12)’ failed. As shown in the figure In fact, only this seed throws the error, and the entire program fails to exit the thread pool properly. I saw that seed_points were generated from getConfigurationSeeds. The positions in getConfiguration define the joint angles of the robotic arm at some positions. Are these configuration positions the end points of the target trajectory? Look forward to your reply and thank you for your help! bro, I got the same problem, have u ever solved this problem? — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.>

Thanks for ur reply! I temporarily solved it by using Drake cpp API to generate iris regions which didn't appear to have this error, then loaded it in Python to the GCS process.