Intelligent Vehicles > Collision Avoidance > Implementation Issues

Printer-friendly version

• Implementation Issues
  • Technical issues 
  • Political issues 
• References

IMPLEMENTATION ISSUES

Several technical, political, and marketing issues have yet to be solved before collision avoidance systems can be safely and effectively mass-deployed. Some of these issues are discussed next.

Technical Issues

The technical challenges for current collision avoidance systems lie in achieving high detection rates given an acceptable false alarm rate, under real-life driving situations. Two specific problems that warrant mentioning here are curved roads and adverse visibility conditions. 
 

The problem with curved roads is that sensors might mistake a car running in the opposite direction, for a hazard on the lane where the CAS equipped vehicle is (see Figure 1).

The problem with adverse visibility conditions (rain, snow, fog, darkness) is different in nature: the issue here is whether the same criterion for collision warning applies under all circumstances. It could be argued that when the preceding vehicle is hardly visible, drivers need to be warned of the presence of a vehicle ahead, perhaps using a worst-case criterion. But when visibility is good, drivers need not be told that there is a car ahead of them; instead a time-to-collision criterion may work best. The results of simulator experiments by Janssen and Nilsson (1991) showed that a CAS based on a four-second time-to-collision criterion and a gas pedal counterforce warning device should not be greatly affected by external environmental conditions, since it achieved both a reduction in average speeds (desirable when visibility is low), and a reduction in the proportion of short headways. 

Design issues also pose a challenge: in-vehicle crash avoidance devices will have to operate in an environment where other displays compete for the driver's attention. The National Highway Traffic Safety Administration (NHTSA) has been studying driver distraction and workload since 1991. The NHTSA has found that 25% of police reported crashes are due to driver distraction (i.e. attending to tasks other than driving). Researchers at the NHTSA have conducted a variety of research projects to determine methods of measuring driver distraction that have looked specifically at drivers' relationships with in-vehicle technologies. One general finding is that voice-recognition is a viable alternative to visual-manual information entry (i.e. on cell phones and route guidance devices). It is difficult to exaggerate the potential safety hazard that an "intelligent" vehicle might pose, if one considers that it might be equipped with information, navigation, collision avoidance and vision-enhancement systems, in addition to cellular phones and other communication devices and a new generation of entertainment equipment.  Read more about ITS Technologies and driver distraction.

Another problem that requires further study is the issue of behavioral changes which result in poor driving habits. Some of the problems that might arise due to the presence of CAS devices in vehicles are: 

• Headway distribution: an overall compression of the headway distribution may result because drivers get to know how close they can get without being warned. Thus headways slightly higher than the critical headway, which may perhaps be shorter than a safe headway given the driver's reaction time, could become more frequent. 

• Higher speed and higher speed variability: the availability of CAS could induce faster and more irregular driving, both of which increase the risk of collisions. 

• Drowsiness: there is the possibility that the driver of a CAS equipped vehicle may become less attentive of his/her surroundings. 

An evaluation of users of an actual warning system shows that, while the majority of drivers agree that they keep longer headways and drive more carefully, several of them (over 10%) think that it has become unnecessary to pay too much attention to surrounding conditions (see Figure 2).

Political Issues

In addition to the technical problems mentioned above, non-technical problems also hinder the deployment of CAS. Some of these need to be resolved by consensus between the government, manufacturers, and consumer groups, while others are probably best left for the market to decide. 

1. Liability: it appears that the issue of who is responsible in case of an accident is one of the largest obstacles in the deployment of CAS, and eventually, of the automated highway system. 

2. Performance measurement: one fundamental difficulty in evaluating the performance and benefits of CAS is that collisions are rare events. Thus special strategies need to be devised to evaluate their performance. Farber et al. (1995) propose to develop and test CAS in incremental steps, starting with computer simulation studies and laboratory experiments, and moving all the way to large-scale tests on public roads. On the other hand, several vehicle manufacturers are already moving ahead with different versions of intelligent cruise control and collision avoidance systems. In several instances these have been installed in vehicles that log above average annual mileage, for example trucks and buses. This provides a unique opportunity to measure the systems' performance over time. By and large, a concerted effort by government and industry will be required to gather data and evaluate CAS performance. 

3. Standardization: due to the large degree of machine-driver interaction, several operational characteristics of CAS will need to be standardized across manufacturers. These characteristics include, for example, warning thresholds, sensor effective range, area of coverage, type of warning, and type and strength of steering or braking intervention. The overriding motivation for standardization is that lack of consistency across vehicles could produce inappropriate or delayed responses by drivers. Similarly, a frequency range will have to be allocated for use by automotive radar systems, to avoid interference with other communications. 

4. Deployment: the development of CAS is taking place largely in the private sector. It is probable that vehicle manufacturers will attempt to recover their R&D costs directly from the sell of CAS equipped vehicles. However, to the extent that second or third generation systems require some amount of road infrastructure (as would be the case for automated guidance), some scheme of cost allocation would have to be agreed upon between vehicle manufacturers and the federal and state governments, in their role as highway infrastructure providers. In addition, there is also the issue of whether deployment will be mandated, as was the case with seat belts and air bags. As the benefits of CAS are yet to be proved, it is probably too early to conjecture upon this issue. 

It has been argued that the use of collision avoidance systems is unlikely to be widespread without government regulation, given that the average driver's chances of being involved in a collision accident which results in serious injuries are about once in 30 years. The lesson from the deployment of similar safety equipment, such as seat belts and airbags, suggests that this is not the case. This is not to say that customers will purchase CAS at whatever price. At present, rear-end collision warning systems add about $4,000 to the price of a commercial vehicle. Clearly, additional technological development, combined perhaps with insurance premium discounts, are required to bring the price of CAS to marketable levels. 

In order to cope with the issues raised previously, the United States National Highway Traffic Safety Administration has a program for development and eventual deployment of collision avoidance technology, which includes five major elements: 

• An expanding crash-avoidance knowledge database. 

• Development of vital research tools, including the National Advanced Driving Simulator (NADS). 

• Identification of crash avoidance opportunities. 

• Examination of key human factors and system design issues. 

• The development of performance specifications for crash avoidance products and systems. 

• More information on this program can be found at NHTSA's Research and Development web site.

REFERENCES

Farber, E., M. Freedman and L. Tijerina. Reducing Motor Vehicle Crashes through Technology. ITS Quarterly. Vol. 3, no. 1 (Summer 1995).

Janssen, W. and L. Nilsson. An Experimental Evaluation of In-Vehicle Collision Avoidance Systems. In: International Symposium on Automotive Technology & Automation (24th: 1991 : Florence, Italy). 24th ISATA International Symposium on Automotive Technology and Automation, Florence, Italy, 20-24th May 1991.

Croyden, England : Automotive Automation Limited, 1991.

Tijerina, L. Operational and Behavioral Issues in the Comprehensive Evaluation of Lane Change Crash Avoidance Systems. In: Transportation Human Factors. Mahwah, N.J.: Lawrence Erlbaum Associates, 1999.

Authors: Francois Granet and Rosella Picado. Last update: 08/05/03