New physics simulator could provide upgrade to UCO science classes

Publication: The Vista, , University of Central Oklahoma , Edmond, OK

Xander Strickland

Reporter

The automotive physics simulator BeamNG.tech released this May, providing an alternative to real-world testing of self driving cars. This new technology includes an list of sensors and user-friendly, editable scenarios that require little coding to assemble, if any. 

A scenario built in Beam.NG‘s base editor glows blue in a dark landscape. (SCREENSHOT/XANDER STRICKLAND)

The engine is primarily used as an efficient and realistic platform for driver training and crash test scenarios, and is advertised as being especially useful in testing autonomous driving systems. BeamNG is effective in these areas because of the unique mechanics and UI that come together to produce a system that can both realistically model soft-body collisions and deformations (using nodes and beams: nodes, which have mass and can move freely, but are restricted by beams, which act like springs which cannot bend or be twisted, but which can stretch or compress before locking into position after an impact is over) and provide depth of customization to scenarios without needing to code.

BeamNG.tech is housed in BeamNG.drive’s existing physics engine, a leading simulator in accurate crash testing.

Many universities and companies have already produced research using BeamNG.drive and BeamNG.tech, including Audi, HDI Insurance, Acusensus, Berkeley University, and the University of Passau.

The BeamNG.tech website features a page providing academics use of their softbody physics engine and driving simulator by registering via the student or instructor’s professional or academic email address. After registering, BeamNG.tech will issue to the student or instructor an Academic License Agreement (which functions as a non-commercial license for all intents and purposes). This BeamNG.tech academic license will be issued on a short-term basis, but may be eligible for renewal.

BeamNG.tech has a section of their website dedicated to offering university students, professors, and other noncommercial or commercial organizations annual licenses to use their software on a free, renewable basis.

BeamNG.tech could integrate solidly into a number of classes taught at UCO, including strengths of materials, physics and even cybersecurity or software development courses.

In an age of safety testing for self driving cars, the engine provides a look at crash tests that is both realistic and inexpensive. (SCREENSHOT/XANDER STRICKLAND)

BeamNG.tech employs data using several other simulation software platforms, including Simulink, a user-friendly piece of multidomain model simulation powered by MATLAB and requires little to no code.

The UCO department of math and science already uses MATLAB to generate demonstrations of and run experiments for STEM course material. The Simulink Onramp is included in the for-students version of MathWorks, which you can sign into for free using your UCO login.

The Onramp is a self-taught course designed to teach the student how to create, edit, and run simulations on models using Simulink, with block diagrams representing chunks of real code. Connecting these blocks instead of having to type out code helps eliminate human error while developing an applied understanding of how to use Simulink.

This means that BeamNG.tech and Simulink together are far more user-friendly than other top softbody physics-driven crash-test and ADAS test software used in the industry, like ANSYS and Virtual Performance Solution. While these leading softwares are highly accurate, their user interfaces can be counterintuitive, especially in the case of ANSYS. For anyone looking to learn how these systems are used, BeamNG.tech and Simulink’s features are easily accessible and have applications in many curricula, from general physics to advanced engineering technology courses.

AI development or cyber-security classes could also be enriched with this technology, as the primary focus of BeamNG.tech is to help provide an easy-to-use environment for ADAS development via an open-source Python-driven interface which includes an array of sensors (from camera to ultrasonic, radar to Lidar).

These two cars represent just a sample of the frame-bending damage that can be realistically explored in the new physics simulator, offering insight into how the engine might be used in a classroom setting. (SCREENSHOT/XANDER STRICKLAND)