Praxis
Praxis builds robotic systems that replicate human dexterity through teleoperation and machine learning. Its platform combines vision-based tracking, articulated wrist modules, and live motion-capture reconstruction to train robots on complex manipulation tasks. The system is designed for developers and researchers building general-purpose robotic solutions.
- Robotics
- Artificial Intelligence
- Hardware
Funding
Raised to date
$500KRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.
Founders
Founder details are not available yet.
Product
Problem
General-purpose robots still struggle with complex, dexterous manipulation tasks that humans perform intuitively. Training robotic systems to handle diverse, unstructured environments typically requires extensive manual programming or large amounts of task-specific data, creating a bottleneck for developers and researchers building versatile robotic solutions.
Solution
Praxis provides a robotic platform that captures human dexterity and intuition through teleoperation and machine learning. The system integrates vision-based tracking with articulated wrist modules, allowing operators to demonstrate tasks that the robot then learns to replicate. A live motion-capture reconstruction layer streams real-time operator movements to the robot, enabling immediate, intuitive control. This approach reduces the effort required to train robots on novel manipulation tasks, letting developers focus on higher-level behaviors rather than low-level control.
Target Audience
Primary users are robotics developers and researchers building general-purpose manipulation systems who need efficient ways to collect demonstration data and train dexterous policies.
Features
- Articulated wrist modules on both left and right sides for bidirectional manipulation capabilities
- Integration with ZED 2i stereo camera for vision-based hand and motion tracking
- Live motion-capture reconstruction with real-time stream visualization for operator feedback
- Four simultaneous recording streams (REC) to capture multiple facets of demonstrations
- Teleoperation-driven data collection pipeline that captures human demonstrations for downstream machine learning training