Autonomous Vehicles – Recent Trends and Test Methodologies

– Archan Mudwel | Technical Marketing Engineer | National Instruments

With the accelerating growth of technology, the Automotive industry is also undergoing a disruption. There is convergence of technologies from diverse domains coming together to build a “vehicle of tomorrow”. These technologies include software, wireless connectivity, sensors and machine learning algorithms. With the accelerated rise in these new technologies, growth in the emerging markets and the change in customer preferences, the automotive industry is ready to see a disruption yet again.

Recent trends in the Automotive Industry
According to a report by McKinsey, the four disruptive trends that are driven by these new technologies include:
⦁ Shared Mobility
⦁ Autonomous Driving
⦁ Vehicle Electrification
⦁ Connectivity
Today, our cars are empty and parked 90% of the time and with the increasing traffic congestion in cities, people are questioning the need to even own a vehicle of their own. Companies like Ford and Toyota who are looking in invest in car-sharing and mobility services platforms as potential business opportunities.

Figure 1: Disruptive trends in the Automotive Industry

Another major trend in the automotive market is the electrification of vehicles. OEMs are aggressively pursuing electrification with Toyota, General Motors, Volvo and Ford making aggressive plans towards electrification. With India leapfrogging to BS VI and Government of India’s aggressive push to electrify all new vehicles by 2030, auto part manufacturers and carmakers are compelled to draw up early plans for electrification.
The other 2 trends are Autonomous Driving and Connectivity. SAE has defined the various levels of Autonomous Driving all the way from no automation to driver assistance and full automation. Most of the systems used today are helping drivers avoid accidents but in just a few more years they may be eliminating the need for drivers at all.

There are various technologies that drive autonomous vehicles:
⦁ Sensors
Sensors are an integral part of autonomous vehicles.

⦁ Radar: Radars typically work by emitting an electromagnetic pulse which gets reflected from objects and is received back by the receiver. Based on the time difference between the transmitted pulse and the reflection and parameters like Doppler shifts, the range and speed of objects is calculated. The Automotive Radar so far has been a tale of two frequencies- the 24GHz band and the 77GHz band.
⦁ Cameras: Radars are effective at detecting obstacles, whereas cameras on the other hand are good at detecting what those obstacles are. There are multiple cameras mounted on a vehicle to capture a 360-degree view of the surroundings. The Automotive Camera market is expected to grow to $5 billion by 2026 according to a report of Transparency Market Research.
⦁ LIDAR: LIDAR (Light Detection and Ranging) works on the same principle as Radar with the only difference being that instead of using Radio waves, a LIDAR uses visible light for creating exact 3D monochromatic images or maps around the vehicle.

Apart from the sensors mentioned above there are other sensors like ultrasonic proximity sensors, etc. that are extensively used.

⦁ Communication and Connectivity
A “connected vehicle” as the name suggests should be connected with the world around it and should be able to communicate to other vehicles (V2V Communication), infrastructure like traffic lights (V2I Communication) and even with the GPS signals coming from the satellites. Combined with other wireless signals of a typical automotive V2X accelerates technology convergence, where test engineers integrate more I/Os and wireless standards. There are different standards competing today for V2X. IEEE 802.11p (DSRC) and LTE V2X (Cellular V2X) being the top-runners.

⦁ Machine Learning and Artificial Intelligence
Machine Learning and Artificial Intelligence needs to play an important role to achieve fully autonomous vehicles. A lot of research is going on to develop deep learning algorithms including neural networks. This kind of inductive learning leads to evolving software operation that is challenging to test.

Test Challenges & Methodologies
Autonomous vehicles pose several test challenges due to regulatory uncertainity, new technologies and their integration.

National Instruments offers a flexible platform based on the open PXI standard that can be moulded to build test systems that can scale up as the test requirements change. This kind of a modular platform allows flexibility and scalability of these test systems.

Let us take examples of some of these technologies and the test strategies for them.
⦁ Radar: There are 2 different tests that are performed on a radar subsystem.
o Radar Sensor Characterization
o Target Emulation or Radar Echo Simulator
NI Vehicle Radar Test System (VRTS) provides automated radar measurement and obstacle simulation capabilities for 76-81 GHz vehicular radar systems. The VRTS’s flexible obstacle generation capability allows engineers to test the embedded software of radar and other ADAS systems through simulation of a wide range of generated scenarios. The system works on the 76-81GHz frequency range and can emulate obstacles upto a minimum range of 4m which a range resolution of 10 cm.
In addition, the combination of high-performance mmWave Radio Heads and the PXI Vector Signal Transceiver (VST) also allows engineers to conduct precision RF measurements for beam characterization and testing.

Major OEMs like Audi have been using the NI Radar Test Systems to test their radar subsystems.
“With the PXI VST, the combination of wide bandwidth and low-latency software allowed us to discover an automotive radar sensor like never before—even allowing us to identify critical bugs in our radar module that we could not detect before.”
⦁ Niels Koch, Audi
You can find the complete application note here.

⦁ Cameras
Cameras help to detect details about the obstacle. Camera sub-systems are testing in 2 specific ways:
o Scene Replay in front of camera assembly.
o Raw Video Frame Simulation with noise and error injection.
Companies like Valeo use the NI platform to maximize test reuse for Vehicle Camera systems meant for Automated Parking.
“Using the PXI platform, we get the flexibility we need, while still ensuring we can reuse all the common elements such as the DAQ functionality and HIL environment.”
-Derek O’Dea, Valeo.

⦁ Sensor Fusion
Sensor fusion is a growing trend where multiple sensors are fused together and their signals are given collectively to the ECU to make better decisions. The systems required to test these need to generate precisely synchronized sensor signals which can be achieved using the timing and synchronization lines on the PXI bus.

⦁ Communication and Connectivity
A typical vehicle contains various wireless standards like radio, navigation, broadcast, cellular communication and connectivity. NI’s RF measurement and test products include Vector Signal Generators, Analyzers and Transceivers to test these wireless standards.
The Second-Generation Vector Signal Transceiver from NI (NI 5840) offers an instantaneous bandwidth of 1 GHz and works up to 6 GHz of RF frequency and can generate and analyze wideband RF signals.

Companies like Harman have used NI platforms for test of wireless standards for infotainment units. “We tested multiple wireless technologies ranging from Bluetooth to WiFi to GPS and cellular all with the same equipment using the NI Wireless Test System which helped us significantly reduce test time and the time it took to get our test systems up and running.”
– Markus Krauss, HARMAN/Becker Automotive Systems GmbH
Please find the full application note here:

Conclusion:
Autonomous vehicles are the future of the automotive industry and the industry is gearing up to move from partial automation or driver assistance systems to full automation. There is a lot of new technology in terms of advanced sensors, sensor fusion and machine learning that is fueling this trend. The test systems that test these technologies need to be flexible and scalable to accommodate new requirements and ensure that these systems are reliable before they enter the roads.