Robot arms

JetMax

A Jetson Nano robot arm with an end camera, ROS and deep learning; Standard and Advanced kits add end tools.

NVIDIA Jetson
Suggested for university

JetMax is an AI robot arm powered by an NVIDIA Jetson Nano. It is built on ROS and programmed in Python, with an HD wide-angle camera at the end of the arm and smart bus servos in the joints. With OpenCV it sorts colour blocks, palletises and tracks targets, and with deep learning models such as YOLO it recognises waste cards, letters and numbers for waste sorting, spelling words and simple sums. It can be controlled from PC software, an iOS or Android app, a wireless handle or a mouse. The Starter Kit works with the suction nozzle; the Standard Kit adds a wireless handle, grippers, an electromagnet and a pen, and the Advanced Kit adds sensor and display modules. The same listing offers a mecanum wheel chassis and an electric sliding rail as add-ons.

Zeta42 is an authorised Hiwonder partner in the Middle East.

Configurations

Choose how it ships.

Model: Starter Kit · Standard Kit · Advanced Kit · Vehicle Chassis · Sliding Rail

What it does

Built for.

Changeable end tools
The end of the arm takes an air pump suction nozzle, a gripper, an electromagnet, or a pen holder and pen. The Starter Kit has the suction nozzle; the Standard and Advanced kits add the others.
A camera on the arm
The end camera mounts vertically or horizontally, for a top-down or a head-on view. OpenCV finds and locates colour blocks for sorting, palletising and target tracking.
Deep learning
Models trained with networks such as GoogLeNet and YOLO let JetMax recognise waste cards, letter cards and number cards, then sort the waste, spell out words or lay out the answer to a sum.
ROS, simulation and kinematics
JetMax is built on ROS with a URDF model and Gazebo simulation. An inverse kinematics algorithm and trajectory planning move it accurately along a programmed path.
Smart bus servos
Three HTS-35H serial bus servos with 35 kg torque and all-metal gears drive the arm and read back angle, voltage and temperature. The brackets are anodised metal, and the controller sits in a metal shell.
Chassis and sliding rail add-ons
On the mecanum wheel chassis, JetMax drives from the app or a wireless handle, follows a line to sort AprilTag blocks, tracks targets and follows fingertip gestures through MediaPipe. On the electric sliding rail, its controller drives the rail's stepper motor over I2C to sort and stack across a longer distance.

Where it is used

Applications.

University courses in ROS, robot kinematics and machine vision
Training and deploying deep learning models on a Jetson Nano
Sorting and palletising exercises, plus gripping and drawing with the extra end tools (Standard and Advanced Kit)
Model production lines on the sliding rail, or mobile picking on the chassis

Videos

See it run.

Specifications

The full sheet.

Robot arm size (suction nozzle version): 475 × 159 × 251 mm
Robot arm weight (suction nozzle version): 1.6kg
Material: Metal and carbon fibre
End effector: Suction nozzle (all kits); electromagnet, small gripper and pen (Standard and Advanced Kit)
DOF: 4+1
Payload: 450g
Power supply: 12V 5A DC power adapter
Hardware: Jetson Nano board and expansion board
Servo type: HTS-35H serial bus servo / LFD-01M anti-blocking servo
Software: PC software + iOS / Android app
Control method: Computer / phone / wireless handle / mouse control (wireless handle: Standard and Advanced Kit)
Programming language: Python / C / C++ / JavaScript
Communication: USB / Wi-Fi / Ethernet
Storage: 32GB TF card
Package size (Standard Kit): 39 × 30 × 19 cm
Package weight (Standard Kit): About 4.3kg