ProtoTwin

ProtoTwin connects to Beckhoff Automation TwinCAT over ADS with extremely low latency. Ideal for demanding motion control applications. Combined with Onshape by PTC and ProtoTwin's CAD-Sync, you have a powerful workflow for Controls Testing and Virtual Commissioning.

💬 "It works on my machine..." The culprit is usually non-determinism. It causes a range of issues: - A bug appears only occasionally, or only on a particular machine. - A control logic change looks like it fixed an issue, but the next simulation run behaves differently. - Two engineers run the same model and get different results. ProtoTwin is completely deterministic by default, even across different browsers, operating systems and hardware. This makes debugging issues and collaboration much easier. If your simulation can’t repeat a test, it can’t validate one!

Vibratory Step Feeder and Segment Feeder Simulation Physics simulation of a vibratory step feeder and segment feeder, created in ProtoTwin Simulate. These models demonstrate the capability of ProtoTwin to simulate industrial machinery involving small and fast moving parts with high fidelity. The bolts used here are 8mm and 3mm diameter respectively.

You can now easily modify the mass distribution for a rigid body. This is particularly useful in cases where collision geometry isn't used, or where the density is non-uniform. You can also scale up the inertia tensor for improved stability with larger timesteps.

Generative AI can help with repetitive tasks when creating digital twins. In this video, our integrated AI agent generates a correctly dimensioned physics-enabled euro pallet with 3 layers of boxes and slip sheets in seconds. This is made possible through a custom harness, which allows the agent to modify the model by executing standalone scripts, rather than making hundreds of tool calls. Try it out today!

Can LLMs program industrial robots? Watch Torq, our integrated AI agent, one-shot program a UFACTORY xArm6 robot to pick and stack parts inside a digital twin. Torq accomplishes the complete task, including: ✅ Adding the robot controller component and configuring the TCP ✅ Identifying the suction gripper and sensor in the model ✅ Programming all the instructions ✅ Wiring up the sensor and suction gripper #IndustrialAutomation #Robotics #DigitalTwin #VirtualCommissioning #AI #RobotProgramming #Simulation #ProtoTwin

Robot solver performance has been improved 7X over the previous version. ProtoTwin can now numerically solve the Inverse Kinematics for 6-axis industrial robots in as little as 10 microseconds.

Starting with V1.1, you can attach images and PDFs to your prompts. This makes it trivial to create custom components for specific field devices. Torq can even create visual representations from images inside of PDFs. In this video, we show Torq creating a custom component for the Sick DT35 distance sensor after being provided the 60-page operating manual. Torq is able to identify all the IO and recreate the sensor inside of ProtoTwin.

Torq is our fully integrated AI assistant for building dynamic digital twins of industrial machines and robots. It can read and modify your model, including editing scripts, creating new entities, and updating existing ones within the simulation environment. This new agentic workflow lowers the barrier to entry for automation engineers working with physics-based simulation.

Digital Twin of a Universal Robotics UR10 Robot Assembling a Magnetic Brake The digital twin was created in ProtoTwin Simulate. We rendered this video using ProtoTwin Radiant. The simulation is entirely physics driven. The robot was programmed using the integrated robot controller component, which solves the inverse kinematics automatically. The CAD for this machine was created in Onshape. ProtoTwin makes it easy to import static CAD and turn it into a dynamic physics-based simulation. This is achieved by adding different components to various parts of the CAD. This model makes significant use of custom physics colliders, dynamic joints and motors, the indicator light component, the robot controller component, the robot TCP component, and custom scripted components. The software allows you to create your own components using TypeScript inside our IDE, which allows you to extend the capabilities of ProtoTwin. The UR10 robot was programmed through the UI with the robot controller component. The movements responsible for placing the magnetic brake in the tray are a scripted robot program created by our AI assistant Torq. The robot controller component is a general solver, which is capable of solving the inverse kinematics of any industrial robot. It is now also possible to program robots with our AI assistant, which makes use of the robot controller component API. ProtoTwin Connect also supports a Python API, allowing you to connect the simulation to Python and control robots with machine learning. ProtoTwin Simulate runs inside your web browser. For more information about the software, please visit our website: https://prototwin.com/ #DigitalTwin #Robotics #Simulation #Manufacturing