Your first two hours
Two hours from an assembled, wired robot to a working VLA example: check wiring → power on → basic teleoperation → deploy and run the project.
Robot assembly and initial wiring are completed before this session and are outside the two-hour scope. Inference software setup and project deployment belong to step 4. The current example is pi05 candy delivery for a compatible dual-arm EDU configuration.
Check your configuration
Four onboarding steps
Check wiring before power-on
10 minStart with an assembled, wired robot. Check power and communication connections for your configuration.
Open guidePower on
15 minFollow the matching power-on sequence; confirm status lights and controller connection.
Open guideTry basic teleoperation
10 minConnect VR, try slow follow motion and gripper control, then return to idle for deployment.
Open guideDeploy and run a simple VLA project
85 minUse the published pi05 candy-delivery weights: set up inference, configure the robot, start server/client and complete a delivery trial.
Open guideWhat counts as complete?
Complete wiring checks, power-on and basic teleoperation, then deploy the VLA example and verify a model-driven candy delivery. A running server alone is not enough. Use published weights; collecting a dataset and training a model are follow-up work. Stop the inference client after the trial and follow the safe return/shutdown procedure.
Robot not assembled yet? Complete installation before starting the clock.