This is a simple example of how to use RobotOps.
Obtain the RobotOps CLI and install it.
Optionally enable tab completion for the RobotOps CLI by adding the following to your .bashrc or .zshrc:
eval "$(robotops completion bash)"To validate the configuration file locally, run the following command:
robotops verify -l schemas/config customers/mxl-shipping/autoFork1.jsonThe output should be:
Status: VALIDNow try validating an invalid configuration file:
robotops verify -l schemas/config customers/mxl-shipping/autoFork2.jsonyou should see:
Status: INVALID
Details:
#config.vision.workflows.0: 2 errors in empty disjunction:
#config.vision.workflows.0.model: conflicting values "tensorflowV1" and "tensorFlowV1":
7:18
./schemas/config/components/vision.cue:12:9
./schemas/config/components/vision.cue:23:12
./schemas/config/components/vision.cue:27:17
./schemas/config/components/vision.cue:27:20
./schemas/config/config.cue:6:10
#config.vision.workflows.0.type: conflicting values "Calibration" and "ObstacleDetection":
6:17
./schemas/config/components/vision.cue:18:8
./schemas/config/components/vision.cue:23:41
./schemas/config/components/vision.cue:27:17
./schemas/config/components/vision.cue:27:20
./schemas/config/config.cue:6:10notice that both errors complain about config.vision.workflows.0.type and refer to the vision.cue file.
In vision.cue the vision block is described as:
#Workflow: #ObstacleDetectionWorkflow | #CalibrationWorkflow
#Vision: {
sensor: "ifm" | "realsense"
workflows: [...#Workflow]
}This means "vision" should have a list of workflows, and each workflow should be either a ObstacleDetectionWorkflow or a CalibrationWorkflow.
The error above is explaining that the first workflow in the list is invalid because doesn't match the expected format of either ObstacleDetectionWorkflow or CalibrationWorkflow.
The definition for the ObstacleDetectionWorkflow is:
#ObstacleDetectionWorkflow: #WorkflowProto & {
type: "ObstacleDetection"
model: "tensorflowV1"
filters: #WorkflowFilters
velocity_factor: float | *0.5
}Notice that model should be tensorflowV1, not tensorFlowV1. Fix that and re-run the validation command:
Status: VALIDBrowse to https://app.robotops.dev (or your own self-hosted instance) and login with your Google account. Create a new organization.
Click API Keys and create a new key. Copy the key and save it in a safe place. You will not be able to see it again.
In your shell, export the key as ROBOTOPS_API_KEY:
export ROBOTOPS_API_KEY=<orgname>:<api-key>robotops schema upload -p schemas/config -v v0.1.0
Now you can validate the configuration file remotely:
robotops verify -v v0.1.0 customers/mxl-shipping/autoFork1.json
# Sending HTTP POST request https://validator.robotops.dev/organization/autofork/schema/validate ...
# Response status: 200 OK
# Status: VALID# This will make a new device called <device-name> if it doesn't exist already
robotops device push-config path/to/config.json <device-name>
robotops device pull-config <device-name>