Problem with the head assembly (Stretch RE1)

Hey all,

I have a stretch RE1 that is having communication problems with the two dynamixels on the head assembly.

It stops responding / is not detected (ran the reboot script) quite arbitrarily. Can anyone give pointers on how to debug this problem? I suspect its a loose connection, but not sure if we should open up the assembly or if it is safe to do so.


@flyingMango, We are sorry you are facing this issue. To get more context on the issue is the dynamixels in the head is always lost/not-responding or do they return back immediately after running the reboot script ( /dev/hello-dynamixel-head)?
And also while this issue happens, do you see the status led of the dynamixel servos blinking red behind the head camera assembly?

They seem to never be responding. When the head assembly is adjusted slightly they seem to work for a short bit before they go back to not responding.

Also where can I find the status LED? I don’t see any such led on the stretch.

Hi Praveen,

If the dynamixel reboot script doesn’t bring the servos back, but physically backdriving the head joints around can make them work again, then it sounds like your suspicion of a loose connection or failing cable is likely correct. If you ping us at we can share a guide detailing how to investigate this further, and how to change the cable out if it has been damaged.

Hi! just an update - still some random behaviour, but can you help me on what this means:

[WARNING] [head]: Device head generating 32.172378 errors per minute
[WARNING] [head]: Device head generating 57.525958 errors per minute
[WARNING] [head]: Device head generating 26.546390 errors per minute

Hi @flyingMango, The warning messages you see are thrown by the DynamixelCommErrorStats class which keeps track of the communication errors with the Dynamixel servos. And it looks like your head dynamixel group is abnormally having a very high number of communication errors raised continuously. So how does the head motions look? Did you find anything abnormal?
If you ping us at, we can share a guide detailing how to investigate this further to check the head assembly cabling as the first step.

The motion is pretty smooth and looks normal, but the errors are thrown even when a simple motion command is issued via stretch body. I have sent an email to!

Hey team!

The robots head is no longer being detected and is not responding at all :frowning:

I am trying to recalibrate the robot, but doing anything seems impossible because it is not working. Requesting your support as soon as possible!

For anyone else who comes across this thread, we solved this issue in a support call by removing the robot’s head shell and dusting out the cables to the motors in the head. If you’re seeing ping errors to the head or wrist joints when running, send us a email and we can help check for damage to the wires and get it working again.

1 Like