Considering autonomous vehicles as physical prototypes of human-safety drivers will have its pros and cons. Analysts continue to challenge this idea, and they suggest that you need billions of miles of road testing to prove that these vehicles are indeed safe under driving scenarios.
MIT’s Computer Science and Artificial Intelligence Lab (CSAIL) team designed a new technology called the Ground Penetrating Radar (GPR) to add this significantly adds to the autonomous vehicle’s advantages.
Regarding their background, the MIT Lincoln Laboratory is an establishment that aims to conduct research and to develop advanced technologies in support of national security. Compared to several national R & D laboratories, Massachusetts Institute of Technology Laboratories emphasizes on building operational models of the systems or blueprints that they design.
So, the CSAIL team devised a GPR to help autonomous vehicles somehow “see” by creating a map of what’s below the road’s surface. The map created by the GPR comes from data measured from a combination of soil, rocks, and roots coming from sent electromagnetic pulses in the ground. This feature is unlike the specifications of previously used LIDAR sensors and cameras.
Teddy Ort: Specs of the New LGPRs and Its Comparison
The CSAIL’s new method, the Localizing Ground Penetrating Radar (LGPR), is a type of GPR designed by MIT Laboratory that has its benefits:
- won’t matter if snow covers the road or if fogs block visibility
- helps self-driving cars to navigate in bad weather
- uses a map to determine its location
- margin of error during bad weather is within inches of it during clear weather
- accurate maps for a more extended time since subsurface features don’t change often
These features are entirely different from the previous devices used in figuring out the self-driving vehicles’ location on the road. LIDAR sensors and cameras do not work well in bad weather conditions, such as:
- unpleasantly wet and cold weather
- lighting conditions
- snow-covered signs
These conditions significantly affect the accuracy and stability of the devices.
Teddy Ort, a CSAIL Ph.D. student, added that LGPR could quantify the specific elements in the dirt, which compares it to the map created. It exactly knows its location without needing those cameras or lasers.
Limitations and Future Improvements for LGPRs
The CSAIL team admitted the systems’ limitations as follows:
- They tested the course at low speeds on a closed country road. However, the research team believes that they could extend the system to highways and other high-speed areas.
- The system works poorly during rainy conditions when rainwater seeps below the road, underground.
- It is not yet equipped for the street roads.
- It needs to combine with other technology for it to detect hazards on the road.
- It is six feet wide, so it has a bulky appearance.
CSAIL Team assured that they now make all efforts to improve on these limits and LGPR’s mapping techniques. CSAIL’s project earned recognition from the IEEE Robotics and Automation Letters journal, and they opt to publish its paper soon.
Mr. Jaycee De Guzman holds a degree in Computer Science. The machine language is his favorite among the several languages he can fluently speak and write with. As a self-taught computer scientist, he is into computer science, computer engineering, artificial intelligence, game development, space technology, and medical technology. He is also an entrepreneur with businesses in several niches such as, but not limited to, digital marketing, finance, agriculture, and technology.