Troubleshooting

Instructions for troubleshooting and sending the application logs

If you're facing issues with AMAS please contact our support team at [e-mail].

Gather logs from the Cortex

Open the command prompt on a PC on the same network as the Cortex, then in terminal type:

ssh extend@<Cortex_IP>

Then collect logs by typing

cat /tmp/extend_robokit.log

Please ensure that your device and the client have SSH installed. You can use this command to install SSH on your Linux machine:

sudo apt-get install -y apt-transport-https openssh-server

Gather logs from AMAS

We highly recommend sending the Player.log file with the email so that our experts can diagnose the application's problem and recommend the best possible solution.

Follow these instructions to gather the log file:

1. Close AMAS

2. Open 'Run' Window.

3. Please type in the following command

   "%Appdata%/../LocalLow/Extend Robotics/AMAS"

   and press enter.

4. Run will then open a file explorer window.

5. Copy the file named 'Player.log' and send the file.

Hand tracking stops working (Rift mode)

In some cases, installing or updating other software can overwrite OpenXR registry keys.

If the hand tracking model is missing on the Rift version, verify that the ApiLayers folder exists in the following registry path using Registry Editor:

Computer\HKEY_LOCAL_MACHINE\SOFTWARE\Khronos\OpenXR\1\ApiLayers

How to check:

1. Open Registry Editor (Win + R → type regedit → press Enter).

2. Navigate to the path above.

3. Check whether the ApiLayers key exists under OpenXR\1.

Notes:

• This issue typically occurs after installing VR-related software that modifies OpenXR settings.

• Missing or altered OpenXR API layers can prevent hand tracking from loading correctly.

How to resolve

• Removing the ApiLayers registry key may resolve the issue.

• Important: Export the registry (or at least the OpenXR key) before making any changes, as a safety precaution.

Why can't my device connect to the internet?

Some Linux-embedded computers have operating system bugs that limit support for specific Wi-Fi devices. To resolve the issue, try to SSH into the device to activate the Wi-Fi driver.

After reboot, open the command prompt on a PC on the same network as the RoboKit, then in terminal type:

ssh extend@<Cortex_IP>

Please ensure that your device and the client have SSH installed. You can use this command to install SSH on your Linux machine:

sudo apt-get install -y apt-transport-https openssh-server

Why is my robot control significantly delayed?

Typically, the teleoperation system's round-trip latency is about 100-200 ms + network round-trip latency. The round-trip latency is defined as the time delay from when the user moves the digital twin to when the user observes the movement of the corresponding robot arm as seen while using the VR headset.

If you observe significantly higher latency, you will likely have one of the following problems.

1. Exit your AMAS application and check for one of the two scenarios:

   1. If the robot arm stops the movement immediately, then it is most likely your PC is not powerful enough. Check our recommended PC configuration on this page.

   2. If the robot arm continues to move for a delayed period after closing AMAS, this means there is latency in the network and/or router. Network delays can be introduced using a low-cost switch between the robot and the PC running AMAS. Check for latency between devices on the network.

2. It might also be the case of bandwidth limitation of your network. You can check our recommended network requirements on this page. The typical behaviour is that:

   1. Your robot is moving with minimal latency (robot control command consumes small bandwidth)

   2. The operator observes delayed visual rendering from cameras, or the render of the feed is dropping frames.

3. Sometimes, the delay can also caused by stressed Cortex devices:

   1. Potentially, your Cortex is running out of storage

   2. Potentially, your Cortex has too much computation running in parallel, such as many ZED cameras in neural depth mode.