Getting Started
Create an account, pick a task, and complete your first robot teleoperation session.
This guide walks you through creating an account, finding a task on Arctivar Hub, and completing your first session.
Requirements
- A modern browser (Chrome, Firefox, Safari, or Edge)
- Desktop or laptop computer (Mac, Windows, or Linux)
- No GPU required — MuJoCo, compiled to WebAssembly, runs the simulation locally in your browser
- A wallet for on-chain signing (created automatically via Privy if you don't have one)
Create Your Account
Sign Up
Visit Arctivar Hub and sign up through Privy — you can authenticate with your email, Google account, X (Twitter), or an existing wallet. Privy automatically generates a wallet for you if you don't connect one. This wallet is used for signing your data on-chain; qualifying runs earn Points in your Arctivar account.

Explore the Hub
After logging in, you land on the Hub — the task hall. Your most recent task sits up top in a Pick up where you left off card; hit Continue to jump back in. Below it, all tasks appear as a high-density list showing each task's scenario, lifecycle stage (Pre — Training — Post), difficulty, and remaining slots.
Finding work that suits you takes two controls:
- The Live | All toggle switches between currently open tasks and the full catalog, and the stage menu narrows to Pre-training or Post-training tasks only.
- The Filters panel narrows by difficulty (1 to 5 stars, labelled Beginner through Expert), scenario (kitchen, home, office, workshop, and more), and skill (Pick, Place, Stack, Rotate, and more).
The To do / Done switch tracks which tasks you've already finished.

Start a Task
Click any task to enter it. What happens next depends on the task's stage:
- Pre-training — you teleoperate the robot to complete the task yourself, producing a demonstration. See Pre-training Tasks.
- Post-training — a trained policy attempts the task while you watch, and you take over to correct it when it slips. See Post-training Tasks.
Either way, the simulated environment loads directly in your browser and the interface shows the goal, the steps, and every control you need.
See Your Results
After completing a task:
- Success: a Task Completed popup shows your completion time. The run is saved to your Portfolio history, where its quality score appears once verification finishes.
- Failure: you can retry immediately — failed attempts are not uploaded and don't count against you.
Each successful attempt generates an independent data entry with its own score. Where a task allows repeats, what it is worth to you is the average of those scores. To make a run count on-chain, sign it from your Portfolio: see Sign & Verify.
Beyond the Hub
The top navigation holds three more destinations:
Alliance — tasks built jointly with ecosystem partners. Every Alliance task awards Arctivar Points plus a partner reward. The current Alliance runs a dedicated subnet on the BitRobot Network. See How Tasks Work.

Explore — the platform's public data tracker: total tasks, trajectories, users, and average score, trajectories per day, the by-scenario breakdown, and a searchable index of every task with its score and data count.

Leaderboard — global contributor rankings across the platform.
Controls
You control the robot's gripper in task space — where it moves, how it rotates, and when it grasps. There are three ways to move the arm, and you can mix them freely:
The fastest way to cover distance:
| Action | Effect |
|---|---|
| Hold & drag the blue ring on the arm | Move the gripper along the current view plane |
| Double-click an object | The arm moves to the object into a pre-grasp pose on its own |
| Mouse orbit / pan / zoom | Adjust the camera viewpoint |
Direct control skips the low-information travel across free space, so your effort goes into the moments that matter: choosing the grasp and the fine adjustments around it.
On dual-arm embodiments you drive one arm at a time: press C to hand control to the other arm, and the movement, rotation, and gripper keys then apply to whichever arm is active.
What's Next
- Learn how tasks, stages, and slots work under the hood: How Tasks Work
- Record your first demonstration: Pre-training Tasks
- Correct a trained policy: Post-training Tasks
- Understand how runs are scored and how points work: Points