Preliminary Results Show a Significant Payoff from New Ramp Meters in Downtown Atlanta

(Last updated 6/1/05)

The recent deployment of new ramp meter controls at four busy Interstate onramps in downtown Atlanta appears to be significantly shortening the duration of "rush hour" in that location, according to Mark Demidovich, Assistant State Traffic Operations Engineer for Georgia DOT. While these results are preliminary – the agency will conduct travel-time studies to verify the benefits shortly and plans to fully document the results in a few weeks – Demidovich remains convinced that the benefits are real and will be sustained. "Preliminarily the results have been real promising, and I don’t anticipate them changing for the worse," he says.

The new ramp meters were installed on the southbound entrance ramps to a common section of Interstate 75 and 85 at Williams Street, Ellis Street, Edgewood Avenue, and Freedom Parkway (see Figure 1). Each of these ramps involves two lanes and is heavily used. The ramp meters are operational only during the afternoon peak hours (2-7 p.m.)

Figure 1. Location of the four new ramp meters in downtown Atlanta

The impetus for installing these new ramp meters was a study commissioned by GDOT and conducted by Day-Wilburn and Associates to do a complete analysis of the Atlanta’s heavily congested downtown area. "It’s eternally congested – especially during the afternoon rush hour," Demidovich says. The study’s recommendations included the addition of new ramp meters on four ramps.

Installation for the new ramp meters began in late 2004 and finished in early April 2005. "We did not let a contract," Demidovich says. "We used our own traffic signal installers and found some surplus signal poles." The ramp meters have been operational since April 5, using a fixed-time scheme. "We’re running them at a very fast rate – about 900 vehicles/hour/lane," Demidovich says. "We have all the sensors in place to do actuated operation, and our intent is to eventually go that way," he says.

While the four new ramp meters are currently operating in a fixed-time arrangement, queue detectors are in place so that if each ramp backs up beyond a pre-determined location, the metering rate will go even faster to help avoid traffic spillover into neighboring arterials.

Impressive Results (So Far)

Thus far, after approximately six weeks of ramp meter operation, the results have been impressive. "We’ve noticed a delay in the onset of congestion early in the rush hour and an earlier recovery at the end of the rush hour, about 45 minutes on each end," Demidovich says. That translates into a shortening of the downtown rush hour, which used to run from about 3:30 p.m. to 6:45 p.m., to about 4:15 p.m. to 6 p.m. (GDOT considers its freeways to be "congested" whenever average traffic speeds fall below 40 mph.)

Demidovich is quick to point out that these preliminary figures are an average, and that specific results vary widely by day-of-week. "The better benefits are earlier in the week, and the lesser benefits are later in the week," he says. "For some reason, Monday has huge benefits, Tuesday is a little bit less, all the way until Friday, when we’ve not noticed not a whole lot of change for some reason. We haven’t completed our analysis of why that is, but my gut feeling is that we’re dealing with much higher volumes on Fridays as people are heading out of town for the weekend."

"We’ve also noticed some increases – and you’d expect this – in average speed," Demidovich says. "We wanted to let everything settle in for four to six weeks, which is right about now, before we start doing some floating car runs where we’ll actually measure trip times through the corridor."

Little Downside (So Far)

Demidovich says that the public response so far to the new ramp meters has been surprisingly muted. "Complaint calls, which have been virtually nonexistent, are one of our primary measures [of problems]," Demidovich says. "I think so far we’ve had 4 people call and say that they didn’t appreciate the wait on the ramp, but two people actually called and complimented us, saying that they’re really happy about the improvement to the freeway. So that nets out to just two negative responses. This deployment definitely results in significant changes in traffic patterns, so such a small amount of feedback is almost insignificant."

Thus far, queue times on most of these new metered ramps during afternoon rush hour have averaged less than one minute, with the longest queue (at Freedom Parkway) averaging around 3 – 3 ½ minutes. Because of the ability to increase the metering rate if the queues get excessive, backup onto the neighboring arterials has so far been very minor. (Some has occurred at Edgewood, which has very little storage.) Demidovich says that long on-ramps are a key factor. "Except for one of the four cases, we’re blessed with long ramps, including fly-overs that have significant storage capacity," he says.

Public Outreach Prior to Deployment

GDOT’s extensive public outreach effort prior to deploying these new ramp controls may be one key to the lack of a strong knee-jerk negative reaction on the part of the public. The department held five public outreach meetings, and kept the local media informed about the plans and expected benefits. "We met with both residents and businesses," Demidovich recalls. "We took the media out to the ramp meters, showed them how they worked, and allowed them to come into our facility on the day we turned them on to film the whole thing. We kept everybody involved -- we didn’t just flip a switch and say ‘here they are.’ I think that was a big help."

Figure 2. Screen shot from GDOT video showing rush hour conditions before the new ramp meters were installed.

Figure 3. Screen shot from GDOT video showing simulation of rush hour conditions once the new ramp meters were operational.

A key element of pre-deployment outreach was the creation of a before-and-after video that filmed actual conditions from an 80-foot high camera. Figure 2 shows a screen shot from that video depicting the "before" case. Figure 3 shows a screen shot of a computer-generated simulation of traffic conditions once the ramp meters would be in place (the "after" case). "The cause of the congestion was so obvious, looking at it [before case]," Demidovich says. "Even the average lay person could see that the ramp traffic was causing huge problems. Vehicles were just piling on from the ramp." Local TV stations (including WXIA Channel 11) aired portions of that video on their news coverage and included it for a while on their websites.

Ramp Meters Not Historically Popular in Atlanta

Ramp meters are not new to Atlanta, although their use has been small and relegated to relatively non-critical locations. In 1996, the city installed ramp meters on five lower-volume on-ramps along northbound Interstate 75. (One of the five has since been deactivated.) Demidovich says that the early ramp meters showed only small benefits, in part because they were located in relatively uncongested corridors. "The resulting benefit was small, because if you don’t have a significant problem to begin with, you can’t expect to have huge benefits," he says. "The five early-deployment ramp meters were mainly an opportunity for GDOT to gain some experience in ramp metering, without the risk of souring the public’s impression of meters."

"The new ones downtown were a whole different ballgame with this traffic situation," Demidovich continues. "The ones that we had from 1996 were on very lightly used ramps, whereas the new ones are on very heavily used ramps. We have dual lanes, and had to do both a very careful analysis of spillback onto the arterials and modeling of the benefit we could expect on the freeway."

Many More Ramp Meter Deployments on the Horizon

Interestingly, plans were put in place prior to the latest deployment to install 120 more ramp meters in metropolitan Atlanta, as part of Georgia Governor Sonny Perdue’s "Fast Forward" transportation initiative. Demidovich says that nine upcoming projects over the next 3 years will install those 120 meters, and that the preliminary results from the four new downtown deployments assuage many of his concerns. "I feed a lot more confident now the way these have operated, with good results on the freeway, minimal negative feedback from the ramp users, and minimal impact on the arterials," he says. "We’ve been processing such a high number of cars. That was a huge concern of ours – was it going to mess up the arterials in the area? The answer has been ‘No" -- so far."

-- Jerry Werner

Mark Demidovich can be reached at Mark.Demidovich@dot.state.ga.us