Thermal Cameras Might be the Key to Making Self-driving Cars Safe

Home > News > Content

【Summary】An Oregon based company claims that heat-seeking cameras should be incorporated into the current self-driving car setup to make these vehicles safer.

Original Manish Kharinta    May 05, 2018 10:30 AM PT
Thermal Cameras Might be the Key to Making Self-driving Cars Safe

With the advances in autonomous technology, the automotive industry is paying more attention to the safety concerns that come with self-driving cars. Companies are trying to incorporate different technologies to make their cars safer and comply with a universal standard of safety regulations. Recently after the near-fatal crash of a self-driving Uber vehicle in Arizona, these concerns were further highlighted. 

The company announced that it is suspending its test and will comply with National Transportation Safety Board during the investigation of the crash. The Lidar used by the company was not able to spot Herzberg in the dark. Speculations claim that it might be because of low resolution or technical blind spots that the operating system was not able to detect and therefore make the appropriate decision.

This proved that the existing technology is not enough to make self-driving cars a practical reality. As a solution to this problem, a company called Flir has proposed that heat-seeking cameras should be incorporated into the current self-driving car setup to make these vehicles safer. The head of product for the Oregon based company explains that thermal imaging cameras will eliminate search errors as they are capable of detecting objects which a standard Lidar will miss because of obstructions. 

The application of thermal imaging cameras is not limited to just detecting body temperatures of human beings which would have come useful in avoiding the near-fatal crash in Tempe. Heat-seeking cameras can detect temperature differences as low as 0.1 degrees Fahrenheit. Not only will these cameras be able to detect human bodies but with a well-calibrated database, they will also be able to detect and distinguish between a wide range of objects.

infra 2.jpg

Lidar uses a laser shooting sensor to detect objects. This technology might malfunction if the view is obscured by fog or direct sunlight falling through the lens. As heat-seeking cameras use infrared light to distinguish between objects such obstructions do not affect the functionality of thermal cameras. 

Technology behind thermal cameras has been intensively developed over the years. Heat-seeking cameras can distinguish victims trapped in burning areas, the technology is also used by Air Force to make the missiles to hit the target more accurately and even detect corroded fuses in electronics. At this point of development, heat-seeking technology has become highly refined and makes it a perfect fit for the automotive industry.

If Uber or any other car company incorporates this technology into its self-driving offerings, it won't be the first time infrared cameras have made their way into the automotive industry. Flagship luxury cars like BMW 7-Series have used heat-seeking cameras to detect obstacles like pedestrians and animals in dark areas where the driver might not be able to observe them using plain sight. However, heat-seeking cameras come with their own imperfections. 

The technology is not cost effective however in comparison to Lidars the heat-sensitive cameras are relatively cheaper. The functionality of these cameras also has limitations. Heat-seeking cameras use infrared Rays to detect objects, but these waves are incapable of penetrating glass and it is at this point that these cameras will fail to function. Regardless it is possible that with effective machine learning and instantaneous exchange of data between Lidar and infrared cameras will make self-driving cars substantially safer.

Prev                  Next
Writer's other posts
    Related Content