3 Speed Management
3.2.5 I-40 Variable Speed Limit System – New Mexico
Impetus for the Activity
A total disregard of fixed speed limits combined with unproductive use of freeways and “speed too fast for conditions” crashes sparked FHWA to examine a technological treatment. It was noted that fixed speed limits are only good for one set of conditions, and that they are too low for most conditions, and too high for adverse conditions. It was also noted that since safe speeds depend on road and traffic conditions, speed limits should depend on these factors too if the technology can be made cost effective.
Criteria for System Selection
The chosen system was selected based on a system analysis of hardware and software requirements. The controller was selected based on its cost, temperature range, ports, speed, ease of programming, portability, and memory. The sign subsystem was chosen based on its legibility, reliability, power consumption, and cost (capital and operations/maintenance). The communications subsystem was chosen based on its reliability (likelihood of interference), capacity, cost, and availability of frequencies. The environmental sensors were chosen based on their accuracy, responsiveness, ease of installation, and cost. The traffic sensors were chosen based on their accuracy, reliability, and cost.
Description of the System
The objectives of this VSL system were:
- Increased voluntary compliance with regulatory signs,
- Greater safety,
- Communication of adverse weather conditions,
- Efficient use of highway facilities, and
- Less burdened justice system.
The system was installed in March of 1989 and was dismantled in 1998 when the freeway was widened from three to five lanes. This prototype system was deployed on a busy urban freeway, I-40 in Albuquerque, but was designed for use in any freeway environment (thus its inclusion here). The system covered 6 km, with three roadside stations, each consisting of a pair of loop detectors in each lane, with a variable speed sign and hazard warning sign on the right side of the road.
As input, the system used traffic speed, light level (day versus dark), and precipitation (wet versus dry). The average traffic speed was computed every 10 s and was smoothed (90 percent old plus 10 percent current). These inputs were obtained by the system using loop detectors for speed, a photocell for light level, and a precipitation sensor.
The system used a look-up table to generate the posted speed limit. The limit was based on the smoothed average speed plus a constant based on the environmental conditions. The constants were as shown in Table 1. Negative constants were used to keep the posted speed below the 89 km/h maximum speed limit cap.
Table 1: Environmental condition constants.
 Table 1: Environmental condition constants. |
The posted speed limit was adjusted in 8-km/h increments and was updated every minute. The posted speed limits were enforceable and were made available to drivers through the variable speed displays on the right side of the freeway.
It was expected that deployment of the system would result in better compliance, more uniform speeds, and reduced primary and secondary accidents.
System Performance
Sign legibility and inability to display credible speed limits higher than the national maximum speed limit affected system performance. It was noted that from an engineering standpoint, the system performed satisfactorily, and ran unmanned for 9 years. The flip disk bleached out due to intense sun and the east-west orientation of the road. The sign was not visible in the morning due to the sun, and it was not conspicuous at night due to competing city lights. This latter factor would not be an issue in a rural area.
System Effectiveness
The effectiveness of the system was formally evaluated using the measures of speed of traffic versus posted speed and frequency of crashes. It was noted that the speed limit correctly tracked the speed of traffic and conditions, as intended. However, in the transition from congested flow to uncongested flow, the displayed speed lagged behind the actual speeds for a 3- to 5-min period due the smoothing process. The number of crashes decreased compared to the before period, but this result was not definitive due to varying road work over the evaluation period.
Obstacles and Lessons Learned
A number of obstacles and lessons learned were noted on the questionnaire:
- Because of the National Maximum Speed Limit of 89 km/h, the system was unable to post speed limits higher than 89 km/h;
- Light emitting message signs are needed to ensure visibility and legibility under all conditions;
- Signs are needed on both sides of the road for freeways with three or more lanes;
- Environmental sensors should be placed downstream of the farthest downstream sign, so that its usefulness can be maximized;
- Posted speeds could change up and down every minute when recommended speeds were near the rounding point. It was noted that hysteresis in system logic will reduce or eliminate these rapid changes;
- While smoothing was based on 10-s averages, the system may be able to get away with averaging once a minute;
- Posted speeds need to be transmitted to police cars/dispatchers;
- A distributed system with intelligence at the roadside is the way to go and can be implemented at relatively low cost; and
- Fully autonomous systems respond quickly.