Autonomous Driving and Semiconductors

Increasing levels of reliability associated with semiconductors, coupled with their compactness and low production costs, have greatly encouraged the incorporation of semiconductors into numerous different applications ranging from sensors to energy systems. More recently, industries, particularly those involved in the manufacture of autonomous vehicles and artificial intelligence (AI) devices, are turning to semiconductor design and fabrication as an ideal method to reduce the weight of end products while also enhancing their technological capabilities.

What is Semiconductor Engineering?
Semiconductor engineering is the complex process of designing, testing, verifying, integrating and manufacturing semiconductors. The aim of the semiconductor engineer is to develop a device that can will easily integrate into the manufacturer’s component later on in the design process.

Autonomous Driving and Semiconductors
In reality, the autonomy of vehicles is graded on a scale from level zero to level five, with zero meaning no autonomy and five signifying complete autonomy. Recent technological advancements in autonomous driving have introduced Level 3 autonomous cars, these include cars capable of carrying out all aspects of driving while a driver is physically present within the vehicle. Current developments in this industry is aimed at achieving Level 4, or higher automation vehicles, which do not require a human driver to be physically present.
As expected, there are extensive complexities and challenges associated with the development of such autonomous driving vehicles. Vehicle autonomy depends heavily on the ability of the vehicle to accurately collect feedback from sensors and use the feedback information to take critical decisions in a rapid manner; therefore, the application of semiconductor technology has proven to be crucial in moving the automotive industry forward.

In traditional vehicles, semiconductors are utilized as microcontrollers for controlling emission release and antilock brakes and also as wireless modem chips for internet connection while the car is in motion. The autonomous features of a vehicle which work on the principles of semiconductor technology include the chips used to provide the vehicle with “eyes” in the form of cameras and sensors present on the vehicle’s exterior. One of the most recent key components of semiconductors in autonomous vehicles is ON Semiconductor, which will potentially provide AI cars with an 8.3-megapixel (MP) camera capable of self-driving and advanced driver-assistance systems.

The Six Levels of Vehicle Autonomy
Engaging your vehicle’s automatic pilot in fair weather may be fine, but any driver would always want the option to regain control of the vehicle when needed.

Level Zero: No Automation
This was a period when vehicles had no computers or microchips in them, and in the early days there were no power steering or power brakes. At the level zero, all aspects of the driving are in the hands of the driver.

Level One: Driver Assistance
This automobile includes some in-built capabilities to operate the vehicle. The vehicle may assist the driver with tasks like steering or acceleration/deceleration.

Level Two: Partial Automation
At this level of automation, two or more automated functions work in tandem to relieve the driver of control. An example of such system is a vehicle system with an adaptive cruise control and automatic emergency braking. The driver still remains fully engaged with the driving task, but it can be noticed that there is the gradual transfer of control from man to machine. This system can be referred to an advanced driver assistance system (ADAS).

Level Three: Conditional Automation
This level is highlighted by both the execution of steering, acceleration/deceleration and the monitoring of the driving environment. In levels zero through two, the driver with his eyes does all the monitoring. At level three, the driver will still be required to be physically present, but the vehicle can independently perform all aspects of the driving task under some circumstances. Another term that can be used to qualify this system of AI integration is highly automated vehicle (HAV).

Level Four: High Automation
Level four vehicles will not need a human driver. The vehicle can essentially drive itself, but the driver can monitor the vehicle remotely to intervene and take control as needed. This level of automation means that the car can perform all driving functions “under certain conditions.” The test vehicles currently on the road would fall under this category.

Level Five: Full Automation
At this level, full vehicle autonomous driving is achieved and a completely automated vehicle will be able to perform all driving functions under all conditions. In this situation, humans will only be passengers in the vehicles.

AI Semi-conductor Chips
AI technology has evolved over time and it has led to a significant rise in specialized technologies which are efficient in medical diagnosis, synthetization of new chemicals, facial recognition, the development of autonomous vehicles etc. Even though AI semiconductor chips are not currently as advanced as many manufacturers would have them be, future progress in this field is expected to leverage semiconductor capabilities to improve the performance while reducing the power of these semiconductor chips.
Research has shown that the effect of future advancements in the AI industry will have massive effects on semiconductor engineering processes. Most importantly, AI will revolutionize the way devices are designed and interact with other semiconductor components, as well as provide upgrades on the tools, hardware and software utilized by semiconductor developers.
The benefits associated with integrating AI and semiconductor engineering processes include increased performance and processing rate by utilizing sparser algorithms to alter the accuracy of specific functions, as well as conduct processing and memory read/writes as a matrix. By varying the way in which data is transmitted along the semiconductor chip, AI is expected to create an enormous difference in the amount of data that can be processed and stored by these devices in the near future.

The Future
Total autonomy of machines is already a reality in some industries. Manufacturing plants are known to employ robotic arms and automated processes to complete everyday tasks on assembly lines. Driverless farm machines plow, plant and harvest the fields on high-tech farms. And the piloting of commercial airlines is nearly all done automatically. Releasing the control of our private vehicles is another matter entirely. How fast and how completely that will happen remains to be seen. Most humans would rather be in control – sophisticated algorithms notwithstanding. Why should we be asked to give it up now?