Oli EDU humanoid robot and the Tron1 EDU biped robot

The Oli EDU humanoid robot and the TRON1 EDU biped robot are the two LimX Dynamics research platforms RoboticsSelect carries, and they answer different questions. Oli EDU is a full-size humanoid with arms for whole-body motion and manipulation; TRON1 is a compact, legs-only machine built to study walking, balance, and reinforcement learning on real hardware.

This article explains what separates the two, how each one stays upright, what sensors and compute ship in the box, how you program them, and which one fits a university lab, a developer team, or a robotics course. Both are listed in our LimX Dynamics collection.

Humanoid or biped: what is the difference?

In robotics, form follows the research question. If the goal is to study how a machine interacts with the world, you need a humanoid shape with arms. If the focus is mastering movement, a legs-only platform removes weight and complexity so the work concentrates on balance and gait.

Oli EDU has 31 active degrees of freedom (DoF, the number of independent joint motions), arranged as 6-DoF legs, 7-DoF arms, a 3-DoF waist, and a 2-DoF neck. It stands 165 cm, weighs up to 55 kg with its battery, reaches about 5 km/h, and carries 3 kg per arm with its standard parallel two-finger gripper; a five-finger dexterous hand is optional. TRON1 EDU stands 845 mm and weighs under 20 kg. Its defining feature is a 3-in-1 modular design: Point-Foot, Sole, and Wheeled foot-ends swap onto the same legs, and the robot recognizes the module automatically and adapts its software.

Spec	Oli EDU	TRON1 EDU
Form	Full-size humanoid with arms	Multi-modal biped, legs only
Height and weight	165 cm, up to 55 kg with battery	845 mm, under 20 kg
Degrees of freedom	31 active (excluding end effectors)	Legs only; three interchangeable foot-ends
Speed	About 5 km/h	Under 1 m/s walking; up to 3 m/s wheeled
Payload	3 kg per arm	Up to 10 kg of research equipment
Runtime	About 2 hours, 9,500 mAh quick-swap module	Up to 2 hours, 210.6 Wh quick-swap battery
Onboard compute	Rockchip RK3588 motion controller	12th-gen Intel Core i3, 16 GB RAM, 512 GB

How does a two-legged robot stay upright?

A two-legged machine is never statically stable the way a wheeled base is; it stays up by sensing tilt and correcting with every step. The sensor doing that work is the IMU (inertial measurement unit), a chip that measures acceleration and rotation, much like the balance organ in your inner ear. Oli EDU uses a self-developed 6-axis IMU, and TRON1 pairs its IMU with joint encoders and force-torque sensors at each foot.

The motion controller reads those signals and commands the joint motors to adjust knees, hips, and ankles. Both robots walk out of the box on LimX's built-in controller: Oli EDU ships with a preset motion library, and TRON1 handles four-directional movement, turning, in-place stepping, squatting, height adjustment, and ground-clearance detection from the remote. In Sole and Wheeled modes, TRON1 can stand back up after a fall. In wheeled mode it clears steps up to 20 cm and inclines up to 30 degrees.

The safety side follows from the physics. The emergency stop on the handheld remote cuts joint power at once, which means a standing robot can drop; LimX calls E-stop activation a high-risk operation. LimX asks for a 1 m gap from every person and a 2 m clear radius before activation, requires a tether rope on TRON1 during initial use, and does not recommend either platform for operators under 18.

What sensors and compute ship in the box?

Oli EDU includes head- and chest-mounted depth cameras as standard, and motion control runs on a Rockchip RK3588 with 8 GB RAM. It does not include a dedicated perception computer; that NVIDIA Orin NX module is an Oli Super feature. For adding your own LiDAR or extra cameras, the body provides USB 3.0/3.2 and Gigabit Ethernet ports plus 24 V/5 A and 12 V/5 A auxiliary power outputs, with Wi-Fi 6 and Bluetooth for wireless connection.

TRON1 EDU takes the opposite approach: strong onboard compute, minimal factory sensing. A 12th-generation Intel Core i3 with 16 GB RAM and 512 GB of storage is installed, alongside the IMU and joint encoders. Cameras and lidar are not included on the base model. LimX offers a Sensor Expansion Kit with a LiDAR and depth camera in pre-optimized mounting positions, a Voice Interaction Kit with an NVIDIA NX module and microphone array, and an Arm Expansion Kit for mobile-manipulation work. The chassis supplies 24 V peripheral power (100 W, 200 W peak), USB 3.0, and Gigabit Ethernet, and the wireless remote works from up to 50 m.

How do you program Oli EDU and TRON1?

Both platforms expose control from high-level locomotion commands down to joint level. Oli EDU's modular SDK has Python and C++ bindings, a WebSocket application protocol, and deployment examples for ROS 1 and ROS 2. It exposes a low-level motion API for joint-level control, a high-level motion API for walking velocity and preset motions, and a sensor API for the IMU and cameras. TRON1 EDU's open SDK supports a complete Python workflow, so C++ is not required, and the EDU edition includes secondary-development access, data visualization, data recording and playback, and joint control.

The standard pipeline is simulation first. LimX ships a complete URDF (the file that describes the robot's links and joints) for NVIDIA Isaac Sim, MuJoCo, and Gazebo on both robots, and publishes its SDK, model files, and code examples on GitHub, including a TRON1 reinforcement-learning training and deployment guide based on Isaac Lab. Train a policy in simulation, validate it on the simulated robot, then deploy the same code to hardware. LimX states that most teams achieve robust walking within a semester; uneven-terrain walking, language-conditioned commands, and multi-robot coordination are year-plus or dissertation-scale topics.

Simulation does not capture everything. Real floors, worn joints, and sensor noise differ from the physics model, so a policy that looks perfect on screen may still stumble on the physical robot. That gap is exactly why LimX publishes calibrated models and sim-to-real transfer pipelines, and why the first hardware runs happen on a protective frame or tether.

Batteries, runtime, and lab setup

Oli EDU runs about 2 hours per charge on its 9,500 mAh quick-swap battery module, with a 58.8 V, 10 A charger included; LimX does not publish a charge time, so a spare pack is the practical answer for long sessions. The robot ships in a foldable seated storage posture that one person can set up, and LimX's manual assumes a protective frame: the robot hangs from hooks on its shoulders with at least 15 cm of foot clearance for development work, zero calibration, and the first stand-up.

TRON1 EDU runs up to 2 hours on its 46.8 V, 4.5 Ah (210.6 Wh) battery. One battery and a charging dock are included; a full charge takes about 1.5 hours, or under 1 hour from 20 to 80 percent, and quick swapping is supported. It is rated for -5 to 40 degrees C in favorable conditions and is neither waterproof nor dustproof, so keep it out of humid, sandy, or dusty spaces.

Which LimX platform fits your program?

Choose TRON1 EDU if your research is locomotion, balance, or sim-to-real reinforcement learning. It is the lower-cost starting point, easy to reposition, and its three foot-ends let one robot compare walking, balancing, and rolling control strategies. In our catalog it is listed under quadruped and legged robots, and the quadruped robot buying guide covers where it sits among four-legged platforms.

Choose Oli EDU if you need a humanoid that both walks and manipulates: two-handed interaction, whole-body motion, and depth-camera perception on a human-scale body for embodied-AI coursework and research. LimX positions the pair as a progression: prove a locomotion direction on TRON1, then scale it to full-size Oli hardware without switching SDK ecosystems. Both sit in our $10,000 and up research tier; the humanoid robot buying guide compares them with the other humanoid we carry.

RoboticsSelect is an authorized LimX Dynamics dealer. Both platforms ship through the manufacturer's official channel with the 12-month LimX manufacturer warranty, verified SDK and firmware access, free shipping to the contiguous 48 U.S. states, and support coordinated by our U.S.-based team. Universities and research groups can order by purchase order through our contact page.

The takeaway

Oli EDU and TRON1 EDU tackle different chapters of the same story: one is a human-scale body for whole-body motion and manipulation, the other a focused platform for learning how two legs stay upright. Pick by research question, plan the lab around supervised operation, and start in simulation before the hardware arrives. Read the full LimX Dynamics brand overview for the lineup and ordering details.