| Robotic simulators have long been used to test code and designs before any actual hardware is tested to ensure safety and efficiency. Many current robotics simulators are either closed source (calling into question the fidelity of their simulations) or are very complicated to install and use. There is a need for software that provides good quality simulation as well as being easy to use. Another issue arises when moving code from the simulator to actual hardware. In many cases, the code must be changed drastically to accommodate the final hardware on the robot, which can possibly invalidate aspects of the simulation. This defense describes methods and techniques for developing high fidelity graphical and physical simulation of autonomous robotic vehicles that is simple to use as well as having minimal distinction between simulated hardware, and actual hardware. These techniques and methods were proven by the development of the Simulation Environment for Autonomous Robots (SEAR) described here.;SEAR is a 3-dimensional open source robotics simulator written by Adam Harris in Java that provides high fidelity graphical and physical simulations of user-designed vehicles running user-defined code in user-designed virtual terrain. Multiple simulated sensors are available and include a GPS, triple axis accelerometer, triple axis gyroscope, a compass with declination calculation, LIDAR, and a class of distance sensors that includes RADAR, SONAR, Ultrasonic and infrared. Several of these sensors have been validated against real-world sensors and other simulation software. |
