Autonomous Vehicles – The Way Forward

We live in what may prove to be the most exciting time in the history of electronics and the world – rapid acceptance and expansion of electronics use in virtually every aspect of daily life. As a result, the world around us offers totally new families of products which are more efficient, reliable, interactive and safer products than most of us could have ever imagined. Nowhere is the trend more apparent that in the current transformation of the automobile.

Figure 1: We live in what may prove to be the most exciting time in the history of electronics and the world - Nowhere is the trend more apparent than in the current transformation of the automobile.

During most of our lives, vehicles have evolved from hand crafted classics controlled by cables and mechanical systems to (ultimately) computer controlled self-thinking machines which cater to our safety, comfort and productivity – in that order…… And if we ever wanted to change the priority of what a car delivered the user it could be modified through software. The current state of autonomous cars is one with head winds coming from regulatory and public acceptance issues rather than insurmountable technical problems. Given that the technology is closer than the vast majority of the public realizes, it is worthwhile to step back and look at what autonomous vehicles (AVs) have to offer, identify the headwinds to acceptance, and examine the impact hardware will play in turning headwinds into resounding acceptance.

Mobility

Mobility is the kind of thing that is taken for granted - until it’s gone. Mobility is sometimes lost though a driver’s inability to drive safely, issues of vehicle affordability, accessibility, and numerous other reasons. The case for autonomous vehicles offers solutions to these problems.

• A fully autonomous car could safely transport passengers (or things) anywhere they (or a third party) desires regardless of physical ability.
• Transportation affordability might well be improved through less fuel and time consumption, plus new concepts of transportation sharing, or off-time leasing. Imagine a vehicle that transports family members as a primary task and also earns income for its owners as an autonomous taxi.
• AVs would reduce congestion though advanced route planning and optimization. Furthermore, there's the possibility of convoying or platooning into "transportation trains" to save fuel and provide the flexibility of routes without rails.

But perhaps safety is the most compelling reason behind autonomous vehicles. With approximately 90% of accidents caused by human error, the advantages of intelligent vehicles are immediately apparent. Yet the concern over autonomous vehicle safety is the topic that has the potential to slow acceptance of, or derail the technology altogether.

Headwinds

Autonomous vehicles currently experience headwinds to adoption with respect to public acceptance, regulatory institutions, legal liability, and the reliability of AVs.

Acceptance - It has been said that the autonomous car is the modern day version of a model T getting riders off horses and into car seats. While that statement is hard to accept at first, informal polls of current acceptance appear to be delineated by age and technical proficiency. Acceptance can be assumed to grow as more and more young drivers are exposed to autonomous options (and elderly choose independence with autonomous cars vs. potentially no mobility or dependence on others).

With acceptance nearly certain to expand, liability and regulatory complexity are the most serious remaining issues.

Regulatory – Several vehicle classification questions exist. These range from concerns of consistent autonomous vehicle performance when exposed to human driven vehicles in different road conditions, complex situations and vehicle age. Additional questions arise regarding what regulations are adequate for autonomous vehicles to operate at maximum efficiency yet protect vehicles driven by humans (e.g., speed limits, lane limits, maximum convoy size etc.) Furthermore, what laws must be enacted to deter hacking of autonomous vehicles?

Liability – Answering who is responsible in the case of an accident is an even more complex question. Does the responsibility of an accident fall upon the person sitting in the driver seat, vehicle owner, OEM, software designers or component suppliers? Regardless of where the liability falls, it’s in the best interest of everyone for hardware to be designed to the highest level possible given the state of the art available.

Reliability - The automotive industry has taken exceptional steps to insure reliability in automobile electronics. From an electronics supplier point of view – implementation of AEC Q200 has had enormous payback on electronic reliability and performance in all areas of transportation electronics. Across the board acceptance of International standards such as ISO26262 governing the safety of critical electronics is imperative in autonomous vehicles.

Emerging technology pioneered by leading component suppliers can complement the existing automotive specifications and standards, and result in a new classification for extreme reliability components and design practice. This might range from an attempt to reduce ECU count and implementation cost (thereby reducing the likelihood of EMI between control processors and possibly even hacking entry points) through the use of multicore processors that can be partitioned to perform certain tasks. Once completed the outcome of the tasks can be acted upon within another portion of the processor core. As automotive grade microprocessors get more powerful, more tasks can be loaded upon the microprocessor. This can ultimately improve processing time and reduce design complexity.

Passive components could play a major role in improving system reliability, especially considering that passives dominate the component count on ADAS systems. For example, consider a mid-complexity ECU having 200 capacitors on it. If next generation self-healing capacitors were used in the design, we have suddenly eliminated 200 potential failure points in the circuit.

A second scenario might be the replacement of discrete inductor / capacitor based filters with integrated LC T filters. In a T filter configuration scenario – two inductors and one capacitor could be replaced with a single integrated filter package which is typically in an 0805 case size. The number of components are reduced, which results in fewer possibilities for assembly induced defects - pick and place errors, damage, board flexure induced failures, etc.

Solder process errors are reduced by a lower number of solder joints. Finally, the failure rate of the single filter is equal to or better than the combination of discrete devices.

Five technology trends are dominating passive component evolution in the automotive sector. Each trend has a meaningful impact on board level reliability. However, their impact on overall system reliability and performance is extremely positive.

1. Self- Healing Capacitors – AEC Q200 qualified capacitors now exist that recover from an induced failure in a capacitive state. The capacitors continue operation as a capacitor indefinitely. These devices are an enabling technology for failsafe power supplies and regulators within automobile (and avionic) circuits. AVX OxiCap® NbO Capacitors are self-healing. When subject to voltage spikes or high current surges that can result in increased leakage current and slightly reduced capacitance, OxiCap® capacitors will continue to function as a capacitor.


2. Transient Capable Passives – Miniature, high Q, low value RF capacitors and inductors have never fared well when exposed to ESD and other transient waveforms. A new material and manufacturing system has been developed to create ESD capable inductors, capacitors and filters. These devices have the potential to harden RF links and sensors throughout ADAS designs (e.g., GPS, Radar, and Lidar). The importance of RF transient capable devices is realized, and pulse withstanding integration capacitors have been developed across multiple dielectrics, case sizes, values and voltages. Pulse Capable integration capacitors can be utilized on IR camera, stereo vision and lane departure / keeping modules to improve assembly yields as well as improve system reliability.


Figure 2: AVX TransGuard® Varistors are rated to 150°C, capable of excellent performance in extreme temperatures under the hood in automotive applications.

3. Broad Band Passives - Broadband EMI filters (for example offering -40dB attenuation across a 500 MHz wide RF spectrum) have been designed to reduce EMI within mixed signal modules (RF:Digital, for example) and to eliminate board-to-board EMI concerns. Next, Ultra Broad Band DC blocking capacitors provide maximum DC blocking on future optical circuitry as well as AC coupling high frequency RF drivers.


4. Harsh Environment Passives – a family of RF capacitor and inductors can expand and contract to FR-4 (and other materials). These devices are capable of working in the continuously changing environment of critical safety sensors. Additionally, a family of capacitors that is capable of extreme board flexure has been created for both power and small signal applications.


Figure 3: AVX TransGuard® High Temperature Multi-Layer Varistors are designed for under hood applications. Products have been tested, qualified, and specified to 150ºC. The MLV has EMI/RFI attenuation in the off state, so one can combine circuit protection and the EMI/RFI attenuation in one device.



5. Miniature Power Hold up Capacitors – Concern has been expressed over the effects of power disruption to critical safety modules. A family of miniature SMT capacitors has been created which is rated in both capacitance as well as joule hold up for ease of end design.

Conclusion

Autonomous vehicles are just a step away. Acceptance will continue to grow while regulatory and liability issues are being nailed down. Electronic designs will migrate to the highest quality component and best design practice across all critical ADAS control functions. Several manufacturers in both passive and semiconductor worlds are already making enhanced reliability and functionality automotive grade components to address the extreme reliability requirements of ADAS, and ADAS acceptance is a stepping stone to acceptance of fully autonomous vehicles. Design topology will evolve and become streamlined to ensure that complex ADAS systems cannot be hacked and EMI does not cross interfere with system functionality.

Ron Demcko, writing for Mouser Electronics, is currently an AVX Fellow and manages the TSG team at AVX corporate headquarters in Fountain Inn, SC. His role centers on projects ranging from simulation models for passive components to product support, new product identification, and applied development. Ron holds a B.S.E.E. from Clarkson College of Technology, Potsdam, NY.