Connected vehicles are one of the most challenging parts in automotive, as they require data processing, Ethernet connectivity, antennae and more connections. The only way engineers can do modern designs in wiring these together is through integration between mechanical and electrical software, which may involve electrical analysis and connectivity issues. Can modern EDA help you here?
Walden C. Rhines, chairman and CEO of Mentor Graphics, speaks to Dilin Anand and Jitin David of EFY
Q. What is the biggest new challenge solved by EDA tools for the automotive industry?
A. Complexity has grown so fast that they cannot manage it in any other way. Especially, now there are safety, regulatory and performance requirements that just make it absolutely imperative that they verify their designs for all sorts of conditions. You need to show a paper trail that you have verified the operations over all extremes of temperature, analysed the electrical characteristics thoroughly, anticipated the impact of various failures that could occur and ensured that they were not catastrophic. The only way you can verify these is virtually.
Q. Comparing with challenges from a decade ago, how have automotive electronics engineers’ need for EDA tools evolved?
A. Since introducing the first wire harness product in 1992, it has been a fairly slow evolution. However, the good news for us is that automotive electronics has increased to a level where it is no longer possible to design wiring by hand. You just have to use automated/semi-automated wiring
software—it is very much like the placing problem in printed circuit board (PCB) or integrated circuit. The engineer has constraints due to all those rules that automotive companies accumulated, and then there are mechanical constraints too.
Q. Could you share an instance of the new kind of constraints that these software help tackle?
A. The software may have constraints such as ‘no more than six wires in a bundle in a wet zone of a car,’ and ‘no more than ten wires in the dry zone,’ or it may have a constraint that says, ‘be sure that all the wire bundles can fit through the hole in the door frame,’ all of which you have to verify without physically trying to force a bundle of wires through the hole each time.
Q. How does software help solve this problem?
A. The only way you can do this is through integration between mechanical and electrical software, which may involve problems such as electrical analysis and connectivity issues. Software today automatically creates a bill of materials (BoM) and tells you how many feet of wire would be required amongst other things. It needs to be three-dimensional and be able to compute distances that are not in a simple XY plane.
There might be companies today who do it all push-button, and then there are companies that do it semi-automatically, but no one today can still design a car manually. In fact, we even have motorcycle manufacturers who use this software.
Q. What is one factor that could have the biggest impact on the EDA industry?
A. With most integrated circuits, the progressions of history have been designing them so that at most you could sell one chip per person or one chip per household and so on. With the Internet of Things (IoT), you go from selling millions or billions to selling tens of billions or hundreds of billions.
So when the number of designs that has to be done increases, the total revenue for the semiconductor industry increases. the information that has to be processed from the IoT requires massive resources for computing analysis. The bigger the IoT, the more things that are covered and the more activity for EDA.
Q. What according to you would be the most challenging engineering aspect of IoT and automotive coming together?
A. Connected vehicles are one of the more challenging parts of IoT in automotive. The IoT includes a lot of other things. For example, we are not far from the point where most vehicles would have Internet connectivity. We will be able to download and upload information as we travel, not necessarily to control the vehicle, but to have entertainment, weather inputs and other data that flows back and forth. That is all a part of the IoT. It requires data processing, Ethernet connections and antennae types of things.
Q. We hear a lot of talk on security within the IoT. Recently I read about the hypervisor solution that has a focus on securing critical information and software. Could you elaborate?
A. Nothing is 100 per cent secure. However, there are a wide variety of capabilities built in to the hypervisor and protocol stack to handle security issues. Those are typically part of the specifications of the suppliers or original equipment manufacturers (OEMs), and these requirements are getting more rigorous every time. Security is important to us in more than just embedded software and hypervisor. It is important in the chips as well.
Q. Could you share an example of securing chips?
A. We have customers who are quite concerned that chips manufactured in a foreign country, for example, might have something added into them that the designer did not know about. So they want us to test the chips and verify that it only performs the specific operations that it had been originally designed for, and does not affect the output or use the signal for anything else. We do development programs in this area, we have basic kinds of software encryptions, but I think this is a very fruitful area.
Q. Is IoT and automotive a promising market, or will it be just a fad?
A. There is a large market for IoT in the automotive without even attacking what is the most difficult problem. We already have the first stage in high-end vehicles today; we have forward-looking radar to slow you down if you get too close to the car ahead. That is connectivity seeing the other car, but it is a safety addition that keeps you from running into the car. That is a long way from Google Car that has been driving now for three years without any accident. It can be done if we are willing to spend enough on electronics. There are hundreds of things you can do with IoT to make cars easier to drive, safer and less accident prone, before you make it driverless.