top of page
Seal.webp

Trossen Robotics Makes THE Official Aloha Kits

Aloha.webp
Aloha Logo

A Low-Cost Open-Source Hardware

System For Machine Learning

ML Research Kits

By Trossen Robotics

Our Most Ambitious Project to Date.

Stanford University Logo

A collaboration with Stanford University on complete, low-cost, robotic machine learning kits for researchers, educators, and engineers. 

What Our Kits Provide

With every Aloha Research Kit you're getting a lab in a box. We take care of the hardware so you can start tackling data collection and data model development to power your machine learning and robotic automation projects.

Project Goals

☑ A Robotic Bimanual Teleoperational Research System
☑ Fine-Grain Controls for a Variety of Training Tasks

☑ Low-Cost Research-Grade Hardware

☑ Repairability

☑ Easy to Assemble and Get Started

☑ User-Friendly

☑ Capture Video for Machine Learning and Data Models

Aloha Hardware

Available in Stationary and Mobile Versions

Whether you are focused on tabletop activities,
or navigating the real world, we have a kit for you.

Official Aloha Kit

Trossen Robotics Makes THE Official Aloha Kits

Aloha Stationary

Aloha Stationary

Version 2.1

$29,999.95

+ Shipping and any applicable taxes and tarrifs

Aloha Mobile

Aloha Mobile

Version: 2.0

$36,995.95

+ Shipping and any applicable taxes and tarrifs

Viper Aloha Set

ViperX - Aloha

Follower Arms Set - Version 2.0

$12,799.95

+ Shipping and any applicable taxes and tarrifs

Widow Aloha Set

WidowX - Aloha

Leader Arms Set - Version 2.0

$8,799.95

+ Shipping and any applicable taxes and tarrifs

Teamwork

Aloha Team

A cross-collaborative team from three universities.

Tony and Zipeng are the lead researchers on the Aloha Projects.

Tony Z. Zhao

Tony Z. Zhao

PhD Student, Stanford University

  • GitHub
  • Twitter
  • LinkedIn
ZipengFu

Zipeng Fu

PhD Student, Stanford University

  • GitHub
  • Twitter
  • LinkedIn
Chelsea Finn

Chelsea Finn

Assistant Professor

Stanford University

  • GitHub
  • Twitter
  • LinkedIn
Vikash Kumar

Vikash Kumar

Adjunct Professor

Carnegie Mellon University

  • GitHub
  • Twitter
  • LinkedIn
Sergey Levine

Sergey Levine

Associate Professor

U.C. Berkeley

  • Website
  • Twitter
  • LinkedIn

If you need product support, please visit the product support page.

Frequently Asked Questions (FAQs)

Q: What are the Aloha Kits designed to do?

A: Aloha Kits are designed to gather the data necessary to develop machine learning models by researchers and engineers at your organization. The Aloha Kits can perform mimicry as a playback of your tasks and come with ACT/ACT++, machine learning policy architectures designed by researchers at Stanford, as well as a simple pre-trained “Hello ALOHA” task policy. This serves as a demonstration of the Kit’s capabilities, and as a baseline for further policy and architecture development.

Q: What data do Aloha Kits gather?

A: An Aloha Kit can record many sources of data. Both Stationary and Mobile kits come with 2 Leader Arms and 2 Follower arms and 3-4 Intel RealSense stereo depth cameras. You can derive joint positions and speed from each servo and image data from each camera. For the Mobile, you can also derive linear and angular speeds from the base motor encoders.

Q: What is the data used for?

A: The data gathered using the Aloha Kit hardware is used for feeding into machine learning models, neural networks or language models. These models can then be used for custom robotic automation applications you develop.

Q: Are the Aloha Kits rated for a commercial or production environment?

A: While we design our hardware for long-term repeated use, it is not designed nor intended for production or commercial environments.

bottom of page