How to Make Your Semi-Autonomous Car Safer on the Highway?

With more than 1 million annual road fatalities worldwide, it’s no surprise that safety is on the lips of every organization dealing with these issues.

December 7, 2018 | 5G

The European Commission has announced massive procedures to cut road deaths and serious injuries in half by 2030. The Road to Zero Coalition, managed by the National Safety Council, released a comprehensive report laying out strategies to end roadway deaths in the U.S. by 2050. And as it happens, in our company purpose we state to “create a safe and ultra-connected society with functional plastics”. Reflecting our purpose and how it would help in achieving the aforementioned targets evoked a need to write a word or two about the effect of the antenna cover, i.e., the radome losses on automotive safety.

As a background research, I checked the EURO NCAP website for automated driving tests with emergency breaking and lane change videos. Very impressive! The videos made me think about ways of still increasing the breaking time.

How can we, as safety driven engineers, increase the object detection distance by several meters or the breaking time by seconds?

To put it simply: the sooner the car detects any obstacles in its route, the more time we will have to avoid the collision by either breaking or switching lanes. This, obviously, means that the radar beam should be longer. Makes sense, but how to do it?

Currently there is a consensus that the two-way radome loss should be less than 2 dB, thus one-way should be below 1 dB. For a 79 GHz radar this means roughly 200 meters of beam length depending on the output power. At 120 km/h, this translates to about six seconds.

To elongate the beam, we should increase the power output. However, increasing the power output is a bit of a brute method of fixing the problem as the radars are already starting to interfere with each other. Additionally, it depletes your car’s battery slightly faster.

How about, if instead of increasing the power, we would reduce the radome thickness to half?

This is a very logical step in the optimization process. I think there is already radars rolling out with only a half a wavelength of thickness. This basically cuts the radome losses to about half as can be seen in Figure 1.

Figure 1. One-way radome transmission loss for currently used radome materials. The thickness of the reference material (turquoise line) is 2.04 mm and the reduced thickness material (purple line) is 1.02 mm.

This, although a very good and justified approach, in turn causes some issues in the manufacturing tolerance and mechanical stability of the radome.  The transmitted power at 79 GHz is actually quite sensitive for manufacturing tolerances. Here’s a graph showing the dependency of permittivity and thickness.


Figure 2. Power transmission dependence on radome thickness and permittivity. On the right, the 90 % efficiency region is highlighted.

Thus, cutting the losses by cutting the radome thickness may actually backfire later with the more sensitive manufacturing tolerances. And as we can see, lower permittivity would be good for the manufacturing tolerances and a radome with weaker impact resistance. Optimal would be to have the radome as close to solid air as possible. (Yes, my colleagues in R&D are still working on it...)

What about changing the radome material closer to solid air?

This is the next logical question. Just as an example, here’s a graph (Figure 3) comparing just the effect of the permittivity. As you can see, this helps only a little if the losses are the same for the new material. There’s a clear benefit for achieving the <1 dB loss over the whole spectrum, and better stability for the beam (lower curvature).

Figure 3. Lowering the permittivity reduces the losses slightly. The thickness of the reference (turquoise line) is still 2.04 and the new material (purple line) is 2.37 mm.

The next logical question naturally relates to the losses. As you can see in the Figure 4. below, the low permittivity and ultra-low loss radome can actually be thicker and still have under a fifth of the dB  losses compared to the industry standard. And we still have room for improvement by splitting the radome to half a wavelength.

Figure 4. The reference material (turquoise line) has a thickness of 2.04 mm and the new material (purple line) has a thickness of 2.37 mm.

Coming back to our initial question on the radar performance, this means well over 40 meters longer beam length and well over one second more breaking time at 120 km/h. I think this added safety is something consumers and insurance companies will be willing to welcome with open arms even before politicians will start making ultimatums.

I’m thinking of writing a series about automotive radomes in the near future. Which topic would you like to read about next?

  • Effect of water and moisture to different radome materials
  • Material effect on manufacturing tolerance and production costs
  • What should automotive (radar) manufacturers take into consideration
  • Other topics?

Please let me know in the comments or by email:

PS. And if it was unclear, you can find very nice radome materials from our portfolio.

Victor Heinänen

Sales Manager
M.Sc. (Chem. Eng), B.Sc. (BA & Econ.) My background is in chemical engineering and business development. I aim to understand my customers’ needs holistically to deliver them a best-fit solution. Even though I am in sales, I try to under-promise and over-deliver. Let’s #makehappen!

Leave a Reply

Your email address will not be published. Required fields are marked *