30-05-2023 | By Robin Mitchell
Autonomous vehicles are dependent on advanced sensor systems such as LiDAR and cameras, but damage to their lenses can result in traffic accidents, something which is already introducing challenges. Recently, researchers from Korea have developed a lens with self-healing properties that could help to improve safety in self-driving systems. What challenges do optical scanning systems face, what did the researchers develop, and how could it help improve safety in the future?
What challenges do optical scanning systems face?
In the automotive industry, there are two hot topics that continue to dominate the media; EVs and autonomous driving. EVs have managed to make some significant progress, with many EVs slowly becoming affordable, vehicle ranges being extended, and charging points becoming more accessible. While progress is being made in the area of self-driving, there are still numerous challenges that need to be overcome, including system reliability, object classification, and the issue of responsibility.
To try and improve the reliability of autonomous systems, engineers deploy numerous sensor systems on vehicles, including LiDAR, cameras, SONAR, and even RADAR. While some (such as Elon Musk) believe that vehicles should solely rely on cameras as “that is how people work”, other, more respectable engineers recognise the advantage of having large amounts of data. By having access to all kinds of sensing data, vehicles are not only better able to understand their environment but also able to adapt to difficult driving conditions such as rain and fog, where cameras would otherwise be useless.
When it comes to visual systems, such as LiDAR and cameras, they generally suffer from environmental conditions affecting their performance. However, there are other concerns that face such systems, and one of those is degradation. Cameras and LiDAR both rely on lenses to focus incoming light onto a sensor, and the quality of this lens directly affects performance. For instance, LiDAR systems, which use light in the form of a pulsed laser to measure distances, can be affected by factors such as ambient light, non-reflective surfaces, and atmospheric conditions. Similarly, camera systems can struggle with issues such as low-light conditions, high-contrast scenes, and fast-moving objects. If, for example, a lens were to be damaged by a piece of debris kicked by another vehicle, it is possible that artefacts in the lens will manifest itself into recorded data, thereby affecting performance.
As such, ensuring that lenses have proper maintenance will be important in future self-driving systems and will likely see the need for sensor replacements during vehicle servicing. However, hairline scratches and chips may not be obvious to drivers initially, and it may be difficult for a vehicle to report that its lens has been damaged. Thus, it is perfectly possible for a damaged vehicle to be driven in self-driving mode, presenting a potential traffic danger to other vehicles, pedestrians, and the driver.
Researchers develop self-healing lens for camera and LiDAR systems
Recently, researchers from the Korea Research Institute of Chemical Technology published a paper on a newly developed lens that has the ability to heal from scratches. When in the presence of sunlight, the lens is able to heal multiple scratches, even those that cross each other, within 60 seconds of exposure, thereby presenting camera and LiDAR systems with a major opportunity in extended life and increased reliability. The Korea Research Institute of Chemical Technology (KRICT) is a leading institution in the field of chemical technology. The research team, led by Dr. Kim Jin Chul, Park Young Il, and Jeong Ji-Eun, has a strong track record in developing innovative materials1
Dr. Kim Jin Chul, one of the lead researchers on the project, explained the significance of their discovery: 'This self-healing lens material represents a major breakthrough in autonomous vehicle technology. It has the potential to greatly increase the reliability and safety of self-driving systems.
Most modern lenses used in optical systems are based on solid materials, such as glass and plastic, which both lack self-healing properties, which is why the engineers turned to flexible materials and polymers. Specifically, the engineers took advantage of a thiourethane structure for the lens itself (something which is already being used in the industry) and a transparent photothermal dye. The transparent photothermal dye forms a complex chemical bond with the underlying lens material and, if scratched, undergoes recombination when exposed to light (specifically, infrared light in the 850 to 1100nm range). This process of recombination is detailed in the researchers' paper, published in the ACS Applied Materials & Interfaces journal1.
Utilising a dynamic polymer network and photothermal dye, a self-healing mechanism is employed in the lens material of self-driving cars.
The recombination occurs as the absorption of IR light induces a temperature increase, and this temperature increase allows molecules in the photothermal dye to move around more freely. Surface tension forces on the surface layer of the lens ensure that scratch layers uniformly heal themselves, thereby restoring the shape of the lens. This process of recombination, as detailed in the researchers' paper2, is a key factor in the self-healing properties of the lens.
How could such a lens help future self-driving systems?
The most significant advantage to such a lens is that it could allow future LiDAR and camera systems on vehicles to quickly heal themselves after being scratched. By doing so, the quality of recorded images and data will be significantly better than those which have become damaged, thus allowing for safer operation of self-driving systems. Such a lens will also be critical in tough environments, such as off-roading vehicles, that often experience dust, dirt, debris, and dangerous weather conditions. This technology could have far-reaching implications for the autonomous vehicle industry. By reducing the need for lens replacements, it could lead to significant cost savings for manufacturers and consumers alike. Furthermore, by maintaining the integrity of the lens, it could improve the accuracy of the vehicle's sensor systems, potentially leading to safer self-driving vehicles.
As legislators continue to mull over many ethical questions, such as “Who is responsible for a self-driving car accident?” the development of technologies such as this self-healing lens will help to reduce the chances of accidents in the first place. Furthermore, the introduction of such lenses will help to instil trust and confidence in self-driving systems that are not always able to record sensor damage, providing a safer mode of transport for those in the future. The Automated Vehicle Safety Consortium™ (AVSC), an industry program of SAE Industry Technologies Consortia (SAE ITC®), is also working on developing best practices and standards to ensure the safe deployment of Automated Driving Systems (ADS). They believe that advancements like the self-healing lens can contribute significantly to these safety standards3.