Anatomy of the Vehicle Safety CocoonPart 1: Safety Architectures - Embedded.com

Anatomy of the Vehicle Safety Cocoon
Part 1: Safety Architectures

The field of automotive safety was essentially born in 1911 when Henry Leland replaced the hand crank used to start the vehicle with a new electric starter after his wife's friend was killed trying to start the car. Since then safety glass, bumpers, safety belts, crumple zones, and airbags have all been introduced with the aim of keeping the occupants safe during and after a crash.

Recently, with advances in electronic technologies, automobile engineers are developing new ways to protect occupants and people outside the vehicle. This “safety cocoon” is becoming more effective and encompasses more features. This article and its follow-on will identify the types of technologies being developed, how these systems work, and the social forces pushing the drive for safer vehicles.

Automotive engineers speak of two types of safety—active and passive. Active safety describes any technology applied to avoid accidents. Passive safety refers to protecting vehicle occupants during and after an accident. Although the technology in these active and passive safety systems isn't totally electronic, much of the new developments are.

Active safety
Active safety technology may be as simple as improving tires and suspension systems, or as complex as creating a complete vehicle dynamics and stability system combined with a by-wire system. The figure below illustrates the different classifications of active safety and their authority with respect to the level of driver control. The size of the circles indicates how relatively effective the technology is in avoiding accidents.

During normal operation, the driver is in almost complete control of the vehicle, and the safety systems remain in the background. The first classification of systems—audio and visual warnings—are intended to warn the driver that a potentially unsafe condition exists. The types of warnings that exist, and the complexity required to implement them can be ranked into three levels:

  • Simple: Warnings that are simple to implement include door and tailgate ajar, driver seatbelt unlatched, and outside air temperature decreasing (possible icing condition). Simple warnings are usually implemented with a switch or sensor. If failures of these systems occur, the failure may be annoying to car occupants but not life threatening—as long as occupants still remember to put on their seat belts. Although many of these warnings have been around for a while, the category is still expanding. For example, today the seat-belt reminder is for the front-seat passengers only. However, both European and U.S. regulatory agencies are actively adding requirements for rear-seat passenger warnings as well.
  • Moderate: These include low tire-pressure warnings and back-up aids that sound if something is behind the vehicle when backing up. Implementing these warnings requires slightly more complex electronics and usually multiple sensors. System diagnostics also become more complex, because a failure of the sensors may cause harm to the occupants of the vehicle or others outside the vehicle. These failures must be diagnosed and the driver immediately warned that the system is inoperable and requires immediate repair.
  • Complex: Intricate alerts include lane change and departure warnings, precrash detection and distance monitoring, blind-spot warnings, and real-time automatic route guidance (telematics on steroids). Implementing these warnings usually requires complex sensor systems working with other vehicle or external systems. Because they're constantly working to identify hazardous conditions that need either immediate or timely attention, these systems usually require complex computations using a microcontroller or DSP. Diagnostics become extremely complicated with these warnings, because failure to warn the driver can cause an accident.

    When any of the above warnings occur, the driver is still in control of the vehicle. Many times, the driver will automatically take some action to respond to or correct the warning, such as stop the vehicle, slow down, or undertake evasive action to avoid the accident.

  • The next classification of active safety systems is assistance systems, which help drivers avoid accidents. The most well-known assistance system is the anti-lock braking system (ABS). If the driver attempts to stop the vehicle rapidly, the ABS system overrides the command and keeps the tires from locking up and causing a loss of control. The driver is still in control of most of the vehicle but no longer controls the braking system. Other assistance systems include tire and suspension technology (nonintelligent systems), vehicle stability, traction control, and adaptive cruise control.

    Implementing these systems requires a complex interaction between many sensors and actuators combined through an intelligent device such as a microcontroller or DSP. Interaction among other systems, such as the engine and transmission is usually necessary to optimize control. Similar to complex warning systems, assistance systems require extensive diagnostic algorithms to ensure the system is working. Failure of such systems to work when requested could cause an accident, and activation when not required could also be unsafe.

    The final type of active safety system is the automatic system, which actually takes control of the vehicle from the operator when the system detects an unsafe condition. Due to the extreme nature of these types of systems, none actually exist in automobiles today, although they're more common in aviation. Examples of technologies currently being researched include automatic braking systems, drive-by-wire systems with automatic overrides (such as steering), and collision avoidance (evasive maneuvers using by-wire technologies). As is obvious, these systems will require extremely complex diagnostics and redundant systems to ensure that they work only when required.

    Passive safety
    Passive-safety technology is applied before an imminent collision and during and after the collision to minimize injury to the vehicle occupants and increase their chance of survival. This technology ranges from “dumb” mechanical systems to complex intelligent ones. The next figure shows passive safety classifications with respect to the level of intelligence required and the additional cost to implement the technology. The size of the circle indicates the relative effectiveness of the system.

    The first classification of passive safety is the vehicle design itself. Examples include crumple zones and interior padding. Because they're part of the design of the vehicle, the relative additional cost is lowest. However, their effectiveness is also low—the occupant may still be injured even in a low-speed crash. This technology has been around since the 1950s.

    Next are restraint systems—seatbelts and child seats—which are also low in intelligence and add incremental cost, but are much more effective. The combination of vehicle design and restraint systems has been the single biggest contributor to occupant safety to date. However, the effectiveness of these technologies has essentially reached its limit.

    The third classification of passive safety is active restraints and cushioning. While it may seem like an oxymoron, these systems include intelligence for activation in the event of a collision, increasing the likelihood that the occupant will survive with minimum injuries. This area of passive safety is the focus of much of the current research into safety. Examples include seatbelt pre-tensioners, airbags all around, occupant “classification” systems, and pedestrian protection devices. The latter two technologies will appear in the near future. Occupant classification systems will be required to identify the size and orientation of a seat occupant and adjust the airbag either to deploy, with a different energy level for smaller occupants, or not deploy. Pedestrian protection devices are a more recent trend. These systems will fire protection features localized in the impact area to protect the pedestrian. Regulations are presently being drafted in Europe and Japan to mandate pedestrian protection, and the U.S. is expected to follow. Technology applied would include styling of the front end of the vehicle, along with sensors that raise the hood, so a pedestrian's head does not collapse the structure under impact and make contact with the engine.

    Because of the complexity of detecting a collision and activating the correct active restraints and cushioning, these systems require a much higher level of intelligence. In addition, it may actually be dangerous to the occupant to incorrectly activate one of these devices. Imagine an airbag firing at 70 mph during routine driving. This could actually cause an accident.

    Another concern with adding more airbags is the effect of the controlled explosions that inflate the airbags. If all bags were to inflate simultaneously, the increase in pressure inside a vehicle with closed windows could blow out the windows and harm the occupants. As more airbags are added, the electronics not only have to detect a collision but also from which direction, so that only the proper airbags are inflated.

    Finally, rescue assistance technology comes onto the scene after a collision. The most famous example of this technology is the General Motors OnStar system, which calls the OnStar service if an airbag deploys. Other technologies are in development today that may prove less complex, such as emergency and mayday signals already in use on boats and airplanes.

    The second of Lemieux'articles provides details of the Electronic Enablers critical to active and passive automotive safety.

    Joe Lemieux is manager of Controls and Electronics at Ricardo, Inc. He has developed embedded systems for the automotive and medical industries for over 20 years and is the author of “Programming in the OSEK/VDX Environment” (CMP Books). He would like to acknowledge the assistance of Ricardo, PLC., and in particular Stephen Channon and Peter Miller, who have performed much research in the area of safety electronics technologies and trends. Lemieux can be reached at joe@osekbook.com.

    Leave a Reply

    This site uses Akismet to reduce spam. Learn how your comment data is processed.