PNDbotics Adam
PNDboticsBiomimetic humanoid using deep reinforcement learning for natural gait
About PNDbotics Adam
Adam stands 167 cm tall (U variant with torso/arms on wheeled base also available), weighs 62 kg, with up to 44 degrees of freedom and 340 Nm joint torque. Five-digit robotic hands provide dexterous manipulation. Unlike traditional robots using rigid mathematical modeling (Boston Dynamics approach), Adam employs proprietary deep reinforcement learning to learn locomotion through real-time environmental interaction. Imitation learning and human motion data enhance gait realism. Biomimetic torso and modular actuators ensure durability and adaptability. AI-powered motion simulation significantly reduced development time and costs. Adam famously played keytar at VOYAGEX Music Festival. Open-source SDK, NVIDIA Isaac Gym support.
PROS
- + Advanced Deep Reinforcement Learning for natural gait; Modular hardware with open-source SDK; High joint torque (340 Nm); Fast skill acquisition via VLA model; Designed for stability in dynamic environments
CONS
- - Currently in early customer trial phase/prototype status; High initial cost for prototype variants; Dependency on complex DRL models for core locomotion
Use Cases
Manufacturer
Quick Specs
- Type
- Humanoid
- Market
- Industrial
- Availability
- Pre-order
- Price Range
- Professional ($50K - $200K)