If you’re tired of hearing about the chip shortage, you’re not alone. Electronic engineers around the globe have had to find creative new ways to cope with the deficit, and perhaps none more so than those in the automotive industry.
Modern vehicles have upwards of a thousand electronic components that increase safety, improve performance, and enhance the driving experience. Although the chip shortage has pumped the brakes on automotive manufacturing to a certain extent, the industry continues to drive ahead towards an electric and autonomous finish line.
The Chip Shortage and other Supply-Chain Issues
According to the Semiconductor Industry Association, the auto industry is the third-largest consumer of microchips, and the rate at which it uses semiconductors is growing faster than that of any other industry. The recent chip shortage has slowed auto manufacturing over the past few years, and many experts believe that the semiconductor deficit will continue through 2023 and into 2024, forcing engineers to find inventive ways to design products with limited supplies.
Ideally, an engineer specifies the requirements and then chooses the device most able to handle the task. But as any football coach will tell you, the best ability is availability, so these days, designers are first looking for the chips that are in stock and then making do with less-than-optimal components. With the power and versatility of today's microcontrollers, this isn't as big of a problem as it might seem. Because microchips are small, light, and inexpensive, engineers have gotten into the habit of designing for distributed processing, where dedicated controllers handle each separate task. Overcoming a chip shortage could involve putting a heavier load on fewer microcontrollers and writing the software to compensate.
In other cases, the chip engineers need is available, but not in the most desirable packaging. In this situation, engineers will redesign the PCB to accommodate the available device and retool the manufacturing machines to assemble the board as needed. Engineers may also specify analog devices that don't require microchip control, such as gauges and speedometers.
Another solution is to design the car with all the desired components, but leave certain features out until supplies can be restored. Maybe you've designed a voice-command system but can't secure the voice recognition chips needed to implement it. The car can be manufactured and delivered without that feature and added retroactively when parts become available. This costs the manufacturers money since they have to pay the labor costs of adding the components, but that's a lesser evil than forgoing the profit margin of selling the vehicle in the first place.
These are short-term solutions to a bigger problem, however. While the pandemic, the war in Ukraine, and labor shortages are receiving the brunt of the blame for the microchip deficiency, part of the problem comes from the shortsighted "just-in-time" (JIT) philosophy of not maintaining an inventory of components. It sounds good on paper until something disrupts the flow of materials. Companies are now considering JIT for some components, but not for those most sensitive to supply-chain disruptions. Engineers are working more closely with procurement professionals and market analysts to hedge their bets against future shortages.
Smart Cars
Much of the new technology is designed for electric vehicles (EVs) and autonomous vehicles (AVs), but some “smart car” tech is applicable to internal combustion engine (ICE) cars as well. For example, AVs use radar to detect potential obstacles, but radar is also used for adaptive cruise control and parking assistance. It can even determine vehicle occupancy and measure passenger vital signs.
Some system-on-a-chip (SoC) microcontrollers are designed for applications that need to meet various Automotive Safety Integrity Level (ASIL) requirements, which include safety features such as ABS, airbags, power steering, and engine management. The makers of these devices offer design suites that include evaluation boards, schematics, code libraries with pre-written software modules, real-time operating systems (RTOSs), PCB layouts, design guides, and bills of materials. Most have integrated development environments that allow engineers to visualize data in real-time.
Moving up the intelligence ladder, the industry is seeing more microcontrollers with built-in features for Advanced Driver Assistance Systems (ADAS) and autonomous vehicle control. ADAS includes adaptive cruise control, traction control, stability control, electronic torque vectoring (to improve handling), pedestrian detection, lane departure warnings, blind spot detection, rear cameras, driver monitoring, and parking assistance. These dedicated controllers have multiple execution units with hardware security features such as domain isolation to reduce the spread of malware and various kinds of authentication to prevent spoofing and protect personal data. AI-specific features include 3D stereoscopic image processing, voice recognition, data compression, navigation, and cloud connectivity. Most have EV-related features to optimize motor efficiency, manage battery charging and discharging, and increase vehicle range.
At the 2023 Consumer Electronics Show, Qualcomm introduced Snapdragon Ride Flex, a microcontroller that integrates the company's Digital Cockpit and ADAS compute platforms into a scalable system-on-a-chip. Snapdragon Ride Flex is designed to ease the transition from manually-driven vehicles to fully-autonomous vehicles through the Software Defined Vehicle (SDV) concept, which allows engineers to add features to a vehicle through programming updates rather than hardware modifications. Auto manufacturers are already eyeing this as a way into the Software as a Service (SaaS) model, where consumers pay a monthly or annual maintenance fee in order to receive regular updates. SDV facilitates Continuous Integration/Continuous Deployment (CI/CD) since new code is simulated on a digital twin, thoroughly tested, and deployed upon completion. The vehicles then provide feedback to the digital twin, via the cloud, enabling it to update its models based on real-world performance.
Snapdragon Ride Flex is scalable, which allows engineers to design one system with “the works” and, through software, disable certain advanced features on low-end vehicles. A single hardware platform reduces the number of different parts that the company must stock and simplifies the manufacturing process (this helps to alleviate the chip shortage problem as well). On the consumer side, someone with a low budget can purchase an entry-level smart car with just L1 (driver assistance) features and, as their budget allows, gradually move up the automation scale to L2 (partial automation), L3 (conditional automation), and L4 (high automation) through software upgrades.
Engineers designing autonomous or semi-autonomous vehicles with the Snapdragon platform can utilize the Snapdragon Ride system design kit, which includes sensors, drivers, hardware and software modules, libraries of safety-certified APIs for ADAS functions, and a slew of configuration, diagnostic, tracing, optimization, and reporting tools. The SDK supports Adaptive AUTOSAR (AUTomotive Open System ARchitecture), a standard for automotive computing systems. Again, standardized components and interfaces help deal with the chip shortage.