Building a modern car involves engineering expertise and software expertise. The two go hand in hand, and the importance of either cannot be ignored. This has given software companies the chance to showcase their know-how in making cars safer with technology. One such company is MathWorks, which has been involved in the automobile sector. To know more about the company’s involvement, we spoke to R Vijayalayan, Manager, Automotive industry and Control Design Vertical Application engineering teams at MathWorks India.
Given the fact that AI is playing an increasingly important role in terms of vehicle safety and in-car entertainment, and the fact that most carmakers are aiming at autonomous capabilities, we wanted to know more about MathWorks’ involvement.
How important is AI when building cars, be it ICE vehicles or electric vehicles?
The growing demand for advanced vehicle features and zero-emission vehicles puts automotive engineering teams under increasing pressure to incorporate new technologies at an ever-faster pace. To meet these demands, more and more teams are moving to AI. The rise of AI over the past decade has produced technologies that can be used at each stage of the engineering workflow. For example, machine learning models can be used to mine historical fleet data to inform key decisions at the design stage; they can also be deployed on the vehicle as part of an advanced driver assistance system (ADAS).
Many recent innovations would not be possible without AI techniques. But not all AI techniques are new to automotive engineers. For example, advanced statistical models (machine learning models in today’s parlance) have long been used to characterize test cell data in calibration workflows.
Automotive teams use AI to:
Build innovative features
Machine learning is best known for its role in automated driving, but it has also made its way into powertrain and controls applications. In these applications, machine learning models can be used to provide estimates of hard-to-measure states such as driving style or component wear. These estimates can be used in feedback controllers to improve vehicle performance or efficiency.
Enhance existing products
Using AI to enhance existing products means that teams can rely on the strong foundation that they have built over many years while augmenting their offerings with new technology. For example, a machine learning model might be integrated into a control strategy only in a region where it is known to be more accurate than existing methods. In fact, for testing purposes, it is common to have AI models running alongside established algorithms to capitalize on the advantages of each approach.
Improve development workflows
As products become more complex and timelines shrink, it is increasingly difficult for engineers to explore the complete design space. Computerized models (whether physics based or data based) can significantly reduce the amount of real-world testing that needs to be performed. Although models developed from physics and first principles are often preferred for their white-box nature, AI-based models can be more computationally efficient in large trade-off studies.
Interpret real-world data
Machine learning algorithms recognize patterns in large data sets, making them a natural fit for the ever-increasing amount of real-world driving data that is available. The trends and patterns extracted from this data can be used for evaluating engineering designs, vehicle calibration, infrastructure planning, and developing new products and services.
Perform process improvement and service enhancement
Predictive maintenance and anomaly detection techniques are being adopted by manufacturing groups for early detection of issues with production lines. AI algorithms can provide an early indication of manufacturing quality (reducing future scrap rates) and predict failures in manufacturing equipment before they occur.
How different is the collected data for autonomous vehicles in India compared to other parts of the world?
The data collected in Autonomous Driving is essentially through the onboard sensors namely: Camera, RADAR, LIDAR and GPS, which help the software to understand the environment and make decisions of accelerating, braking, and steering. While the type of information collected in India will be like other parts of the world, the complexity lies in the environment and hence the number of actors in the data. As an example, Indian traffic can have more two and three-wheelers which is unique to our environment, different road signs (or in some cases the lack of it), different road objects such as number of pedestrians, slow moving traffic and cattle in rural roads and so on. These are unique to India which means that the AI and software algorithms used in Automated Driving need more rigour of testing to check for failure in edge cases for operations here.
It is estimated that algorithms must be tested for billions of kilometres; this is another area where simulation will be leveraged to achieve higher confidence of algorithms. With simulation, there comes the need to make it closer to the real world so that we can have robust validation – especially the number of edge cases. In addition, today our prototype vehicles are generating lots of data that engineers want to bring back to their R&D to improve their vehicle designs.
Keeping this in mind, MathWorks has developed a workflow that brings recorded data from autonomous prototype vehicles into simulation using sensor, GPS data and map information. Once you have created a real-world scenario in simulation, limitless variations can be made to these real-world scenes to test edge cases – for example adding a pedestrian running into a lane, sudden breaking of the vehicle in front of the ego car which are typically difficult to achieve in a real world. Such workflows have the dual benefit of customers testing their algorithms for many miles and making the simulation close to the real world.
Can AI help cars communicate with each other or with traffic lights in the future to make roads safer? What more can it help achieve?
Connected cars technology entails communication of the car with the environment. V2V infers vehicle to vehicle communication and a broader V2X infers vehicle to infrastructure. There is an array of possibilities when vehicles can communicate which include safety use cases like keeping speed limit, clearing the lane for an emergency vehicle and commercial use cases such as enabling fleets to communicate with each other for platooning to name a few.
To achieve connected functionalities the vehicle must understand its environment where AI comes into play and communicate, which in India is essentially through telematics, over 2G, 3G and upcoming 5G. In a way connectivity brings communications and the Automotive world together.
MathWorks is looking at connected technologies from two aspects: (1) the connectivity technology (5G, 4G LTE, Wi-Fi, Bluetooth, etc) (2) Web/cloud-based applications leveraging data aggregated from multiple vehicles.
MathWorks provides solutions for developing connectivity technologies such 5G, 4G LTE, Wi-Fi, and Bluetooth ranging from simulation and analysis of network performance and signal strength to the development and testing of modem hardware and software. The data aggregated from connected vehicles enables various fleet data analytics applications such as predictive maintenance, fleet optimization, and driver/ride quality assessment. MathWorks provides a combination of GUI-based and programmatic tools for the easy development of such advanced analytics solutions that can be rapidly deployed to web and cloud platforms.
The future of Automated and Connected is closely tied and MathWorks continues to invest in both areas to accelerate development for our customers.
How far away is India from seeing autonomous cars on-road?
As we all know autonomy is defined by SAE at 5 different levels. The L1 and L2 which are often called as driver assist systems are already being offered and monetized in India. In the short term, driver-assist features — especially in safety functions, such as Emergency braking and comfort functions like park assist — are expected to be increasingly offered in India. Higher levels of autonomy are possible, in more controlled environments, for example: highways, smart cities, tech parks, etc. Once we gain more confidence in controlled environments and are able to define homologation and regulation there can be a runway for Automated Driving (Level 4 and Level 5).
As R&D engineers, many of the design and development frameworks used in driver assist can be leveraged in automated driving as well. As a simulation partner, MathWorks is investing in various areas to strengthen R&D in aspects such as creation of complex scenarios, building sensor models with the required fidelity, creating virtual test benches for closed loop simulations, improving the AI algorithms for perception, improving sensor fusion algorithms, generating code for GPU targets to name a few. As engineers have the right frameworks for development and validation this will pave the way for confidence from an engineering standpoint to offer higher levels of autonomy.
MathWorks has been facilitating Executive Round tables for the last couple of years. These efforts have been bringing Industry experts representing various sections – OEMs, industry associations, startups, and suppliers – to a common platform to discuss challenges, specifically for India and find way forward.