Record and Upload Tool

Before running the recorder app, update the AWS config file on the robot as shown below.


ssh nvidia@<ROBOT IP> mkdir .aws cd .aws touch config echo "[default]" >> config echo "region=us-west-1" >> config

Use the following command to pull and run the Docker container with the data recorder application.


docker run -it --rm --privileged --gpus all --network=host \ -e AWS_CONFIG_FILE=/etc/aws/config -e AWS_SHARED_CREDENTIALS_FILE=/etc/aws/credentials \ --mount type=bind,source=/dev,target=/dev \ --mount type=bind,source=/etc/nova,target=/etc/nova \ --mount type=bind,source=$HOME/.aws,target=/etc/aws \ --mount type=bind,source=/sys/devices,target=/sys/devices \ --mount type=bind,source=/tmp/argus_socket,target=/tmp/argus_socket \ --mount type=bind,source=/sys/bus/iio/devices,target=/sys/bus/iio/devices \ --mount type=bind,source=/media/orin_ssd/recordings,target=/media/orin_ssd/recordings \ --mount type=bind,source=/var/nvidia/nvcam/settings/camera_overrides.isp,target=/var/nvidia/nvcam/settings/camera_overrides.isp \<YOUR_STAGING_AREA>/data_recorder:isaac_2.0-aarch64 \ --param=uploader.s3_uploader/s3_uploader/bucket=<YOUR_S3_BUCKET>/mapping_recordings


Replace <YOUR_S3_BUCKET> with your AWS S3 bucket.


Replace <YOUR_STAGING_AREA> with your assigned NGC Registry staging area.

Open a browser (Chrome or Edge) on a browser-enabled device, such as PC/Laptop or Tablet, that is on the same network as the robot. Open the following address to access the user interface:



The interface will then prompt for a password. The default password is isaacsight. To configure a custom password, append the following line to the run command:



On the browser the Recording GUI will look like the following:


In the Recorder tab, the UI will display sensor information and a record START/STOP button. The input fields can be used to add metadata to the recording.

The metadata fields on the right should be filled before recording starts. This data can be used later to identify the recording. Input your user ID into the Author field and add a Title, Location, and Description for the map. Also, add the following tags:

  • “customer_token:xyz” (where “xyz” is provided by the NVIDIA Solutions team)

  • “map_id:xyz” (where “xyz” is the map ID that this recording is targeted at and provided by the NVIDIA Solutions team)

The Robot Id should default to “Carter”, which is the robot that will be used for data collection. If this value is set to anything else, select “Carter” from the pulldown menu. Ignore the Custom Id field. The Robot Id will be included in the metadata of the recorded pod file, serving as an indicator of the robot that was responsible for recording the pod.

The Next and Previous buttons can be used to cycle through the sensor previews for a quick visual check. The indicators on the UI indicate the health of various components. A red indicator signals some issue with the subsystem.

A recording can be started by hitting the START button. The button text will change to STOP once the recording has started. Hitting the button again will stop the recording and will transition the text back to START.

During recording, the stopwatch will begin, and the disk status indicator will give an estimate of how much recording time is available given the current disk capacity. Note that the recording will stop automatically if the disk reaches capacity.

Use the joystick to move the robot along the desired path for recording.

Once recording is done, it is time to upload the data. As mapping is currently the only use case for map upload, all data being uploaded will be used for map creation.

Before uploading, we recommend using an Ethernet cable to connect the robot to the Internet. While it is possible to perform uploads over WiFi, it is much faster to do so over Ethernet.

To perform upload from the Recorder UI, select the Upload tab at the top. Select all files you wish to upload by checking the checkbox on the left of the desired files. Press the Upload button to upload the selected files, or press Upload All to upload all the files. The Status column on the right will display the upload status of each file.

Files can be deleted from the robot when their upload status indicates 100%.

© Copyright 2018-2023, NVIDIA Corporation. Last updated on Oct 30, 2023.