Altair Aids Systems of Autonomous Driving

By T Murrali:

Altair Engineering, a leading provider of engineering software, offers simulation- led design solutions for product design and development. At the recently-concluded annual Technology Conference, AutoParts Asia caught up with MasatomoAyame, Senior Vice President, APAC and Country Manager, Japan, Altair, and Pavan Kumar, Vice President South Asia and Managing Director, Altair India, to understand their operations and contribution to the automotive industry in India.
On the key drivers of the company, Kumar said, “Fundamentally, what is driving the business of simulation is the time-to-market, cost, quality and reliability. Time-to-market is coming down every day; customer demand for lower price and higher quality is going up. It is an ever-moving mark. Customers can react to this by doing as much as possible in the digital world and as little as possible in the physical world. That would continue to drive our business constantly upward. This is the reason why analysts predict double-digit growth for the analysis business. The CAD geometry is created once but it goes through physics based verification many times.”

Expanding on the company’s strategy of a digital twin, he said, “There are several ways of creating a digital twin of a physical device. One is the mathematical equation, a 1-D way of representing the model; 1-D is basically the mathematical model where there are blocks to represent different subsets showing an input and output. Then there is the 3-D representation, which in our case is the CAE (computer-aided engineering) model of simulation that is full fidelity. Our IOT strategy is utilised to connect the instrumented physical model to both the math and CAE model and we can now establish a constant feedback loop which will help capture real life “as used” data with the virtual model, thus helping us increase the fidelity of various real life scenarios like NVH, crash, durability etc.
Kumar further said, “In normal simulations we are studying the ideal state. We assume there is no manufacturing defect with all conditions and usages as defined by us. Let’s take for example durability. Say the vehicle is designed for 10 years of useful life with regular service intervals. When we actually produce the product, it comes with its own differences. A bad weld somewhere will compromise the whole system; there is no way of knowing how the real product will really perform. But with the advent of sensors, IoT, etc we can measure some key parameters. Let’s say you have some strain gauges on the vehicle that are connected to a device sending information to the IoT cloud. This information comes back in the CAE model. If I can input the strain value and rerun the simulation I can take corrective action as required. This means that you can re-predict the life of the vehicle. It might indicate that you have to stop the vehicle immediately or recall it back to your workshop to fix the problem. Today you have diagnostic tools to predict the life of the vehicle based on real- life monitoring.”
Two things can happen here. One, attend to the vehicle for servicing, etc, and two, take this input to make a better product so that the failure rate is minimised. In a large fleet of trucks if it is possible to collect the right information using GPS, GSM and other devices, and start feeding it into the digital model, then it will result in getting the true digital twin of a real product. Altair can grade the model with real information, he said.

Ayame said, “Most of the OEMs today are trying to understand how to deal with the digital change. They are wondering whether to make vehicles or just run them. The business model is changing. When you move into mobility as a service rather than as a product, you need this information to ensure you make money.”
Asked what edge does Altair bring to the table, Kumar said, “The true edge is the CAE rerun, the ability to handle the simulation rerun and re-predict the life of the vehicle. That is the USP we bring to the table. There are some issues along the way likesgetting too much data that you don’t know what to do with. There are solutions here too; you don’t need to send everything to the cloud. What you need is a combination ofa good edge device coupled with analytics and probably techniques like deep learning or machine learning algorithms. You need to filter information before sending it up. Altair has the full system to do this. We have an IoT platform with AI; we understand HPC, deep learning and CAE. I doubt if there is any other company that can comprehend this so well.”
Altair has successfully integrated the existing elements with IoT to help enhance support to customers so that the service ecosystem is complete. Ayameexplained, “IoT solutions are not standard; each company has a different definition of it especially on the customer side. We have put some of our main IoT solutions, not engineering, on the market. We are looking more at manufacturing, using IoT with sensors. We are a platform and technology provider and to make IoT work, it has to be customer and industry-specific. For example, if a genset manufacturer wants to understand the working conditions, he first needs to know what to monitor. Just having sensors all over is not going to help. You need to really understand what it is that you are monitoring. It calls for domain knowledge and you need to build the solution around it. There are system integrators that do this. We at Altair are not contemplating to be everything for everybody. We are probably looking at remaining on the platform which has many pieces to the puzzle. We are working with some customers on some pieces of the puzzle but we will not be doing all of it. It’s like saying you have all the software tools so why don’t you make your own car; it is just not possible.”
Using the genset example, the product can be deployed in different weather conditions. So how can your customers retrieve data that could be used for their improvement?” Kumar replied, “IoT is a big thing; it can be used everywhere. Where we will eventually converge will be on what we know best, which is the ability to work with automotive companies. We understand how engines work; now we are trying to understand how electrical and control systems work. So we would develop a set of solutions that are specific to this domain.”
Optical solutions utilising simulations made with the help of optics like laser, LED, etc are missing from Altair’s portfolio. However, most of the comprehensive solutions are already in the market. The market is pretty small for optics like, for example, how will the interior look when the lights are switched on or how will the headlamp light spread be. They are looking at it more from the design point of view; how they could be of help to the designers.

Autonomous Driving

All the disruptive technologies coming in today in the auto sector would lead to autonomous driving. When asked if there were any particular initiative taken by Altair to support this, Kumar said, “Autonomous driving is dependent on several pieces of technology working together; Radar, antenna, control systems, etc connected with a compute infrastructure that will do everything online using the cloud to make decisions. We have solutions that will help you predict how an antenna will behave if it is placed on a surface; if there are any gap in communication due to shadow areas, or if radiation from antennas could harm those inside the vehicle. The strength of the signal, its propagation and coverage area are all part of electromagnetics. For control systems we have three tools – compose, activate and embedd – which we have created. 0-D is compose, 1-D is activate, and embed is a fully automated code generator. Then comes the compute infrastructure. Thanks to the fact that or PBS-pro manages all the major supercomputers in the world, making this computer on board work better is something we have done as a job scheduler. In the car you need to take many decisions fast for which you would require the run-time scheduler. The last piece in the puzzle is the motor or actuator that controls all this; we have a solution for it. The various tools that I can think of are, FEKO for antennas, wave propagation studies through Winprop, Motor design through Flux, for compute infrastructure there are two solutions – PBS-pro and Run-time.”
Is there a particular megatrend that drives Altair to bring out solutions for customers? To this Ayame said that earlier solutions were based on a couple of predominant megatrends changing the world like the move to alternative fuels and propulsion systems, 3-D printing, AI, etc. They now focus more on Asia-Pacific, particularly India. India is a significant business in the APAC region; Japan continues to be the big Daddy with India a close second, followed by China, Korea and Australia as well as some of the smaller countries from the ASEAN group.
About the plans of the company, Kumar said, “First thing is to determine the future – short-term (one to two years), mid-term (two to five years) or long-term (five to ten years). As a company we should be able to adapt our tools to handle the changes. In India what is now relevant is BS-VI, for which we have to concentrate on facilities like after-treatment, better combustion and related areas. You can also achieve that by reducing the workload on the engine by bringing down the weight and the drag. If I can reduce the drag on a truck that is going at 70 kmph on the highway, probably I can save 20 percent fuel there itself as I don’t have to run the engine at full capacity.Additionally, lightweighting is something that we do fulltime; that’s where we are completely relevant today. This is the short or near-term.
For the mid-term, in the Indian context, I would say it is the whole push towards alternative propulsion which is electrical or hydrogen- oriented. For electrical it is our acquisition of Flux to support electric motor design. Automotive companies will have to design their own motors to give the right torque characteristics to run the car, with a control system to take care of it. We are producing those tools and techniques as also building up the capability within Altair to help our customers. We now have a team of 30 people in India learning these tool sets; the whole embedded development is happening here with a big crew, all working towards applying this technology in the market.
For the long-term it could be the whole concept of autonomous driving. That is where we are working with FEKO, Winpro and Control Systems. It’s still a vision that may not be reached but we need to be equipped. Altair is probably the only simulation company that has engineering manpower with it. We don’t just sell the software tools; we understand how it is applied. That is why we have a service wing which we never closed; we have 500 service personnel who are essentially automotive engineers who understand how systems adopt. This is applicable across the globe; every office has a mixture of product design specialists, and automation specialists to help avoid duplication of work, with sales and support teams.”
Regarding the update on service and software development, Kumar said it is always growing, with more people being hired; appointment advertisements are continuously coming out from Altair

Leave a Reply

*

This site uses Akismet to reduce spam. Learn how your comment data is processed.