Issue link: https://iconnect007.uberflip.com/i/1293772
44 SMT007 MAGAZINE I OCTOBER 2020 cars through sensors and increased data intake. Those sensors produce a lot of data. You're using computers to make decisions about what the car sees, if not eventually what the car does. There are a lot of architectural changes that are happening in the automobile. As a result, PCBs are going to get more complex. Johnson: While they're getting more complex, it's also true that automotive as an industry is driving PCB manufacturing to do two things at the same time. One is to get two orders of mag- nitude more field reliability and accuracy than we currently produce in the PCB fabrication industry and as much as two orders of magni- tude more volume. Those two things often go in opposite directions. While things get more complex and smaller dimension size to fit these things into a mobile platform and better handle environmental extremes, all of this is going on, and there are a lot of dynamics for the PCB fab- ricators, which circles back to around to why this is a great conversation for us to have. Rice: The second part of what I see as a big driver is the electrification of cars in general and electric drives. All of the electrification drives circuits for motors, which drive the car, as well as the tons of different fans and other things for cooling. Those items are going to drive up a lot of semiconductor, electronic, and PCB content. Andy Shaughnessy: I'm curious about the approach to writing the chapter. A whole group contributed to this automotive chapter. It sounds like you had a lot to tackle, especially to get down into the vehicle to vehicle commu- nications. A lot of it has to do with sensors to make all this work. Rice: Urmi Ray worked with a different co-chair to start this process. I caught the ship while it was moving already, but I know how we got here. Our process was first to look at all the areas in the automobile. There are a lot of reports available that talk about how automation and electrification of cars are going to increase semiconductor content. Then, we identified the areas that will require a lot of semiconduc- tor content and innovation to achieve particu- lar long-term goals that the automotive indus- try wants to tackle. For instance, in the area of autonomy, it's going to require a huge com- puting resource and a data-crunching resource. And it's also going to require a certain level of conductivity and data communication. We tackle those couple areas by looking at the communication processes or roadmaps to understand the differences between computing or mobile applications, or maybe cellphones, versus what automotive is going to require. Of course, we had to delve into not only high-per- formance computing and how to deliver it in an environment that's not in a data center or your office. It's a mobile platform that's exposed to temperature excursions and humidity. It has to be robust and reliable, and it has to last longer than your home computer laptop. We researched those areas, and sensors are an important part. We saw that radar contin- ues to evolve. We didn't have radar in the first revision of our chapter, but the second revi- sion—which we're working on basically as we speak—will have a pretty large section as far as radar systems and how they're evolving. LIDAR is something that's on a future roadmap. We have a work group member from Velodyne, for instance, who is a developer and provider of LiDAR. We get their perspective as a con- tributor to the chapter. On the electrification front, if you look at all the power circuits out in the world, about 50% of them go into cars. The other 50% supports all other applications, including industrial, home appliances, etc. Automotive represents a big power require- ment that is only going to get bigger as the penetration of electric cars goes from 4% of all cars today to the projection that by 2030 fully 50% of all cars will have some type of electric drive in them. That might be a hybrid electric vehicle, but they'll still have an electric drive. We're relying on the power group in the HIR to also provide a lot of insight on things like wide bandgap materials, such as silicon carbide and gallium nitride, that allow power transistors to function faster and more efficiently. We've had to just look at those different areas where we