Office of Operations
21st Century Operations Using 21st Century Technologies

Ramp Metering Presentation

Printable Version [PDF 322KB]
You may need the Adobe® Reader® to view the PDFs on this page.
Contact Information: Neil Spiller at Neil.Spiller@dot.gov


Slide 1

Picture of a red signal indication with adjacent sign "One Car Per Green". Picture of a shoulder-mounted ramp meter setup.

Ramp Metering

The application of control devices to regulate the number of vehicles entering or leaving the freeway, in order to achieve operational objectives.


Slide 2

Brief History

  • 1963: First use — Chicago; Eisenhower Expressway — traffic officers would stand on ramp and release vehicles
  • 1964 — 1967: Detroit and Los Angeles – (although no permanent meters were installed for a number of years thereafter)
  • 1970: Minneapolis area – "fixed time, permanent" (including a bus bypass on some ramps to encourage transit use)
  • 1972: Minneapolis area – the first "coordinated" meters were installed on multiple ramps on facilities
  • 1980's and 90's: advancements towards "traffic responsive, dynamic" meters that would "self-regulate"
  • 2000: Minnesota's public "push-back" against meters
  • Circa 2006: CALTRANS District 7 advanced "System Wide Adaptive Ramp Metering (SWARM)" to control whole freeway corridors automatically

Slide 3

Types of Ramp Metering

  • Stand-alone (i.e., "time of day") 1960-70
    • Often manually operated (switch on, switch off) or simple "traffic cop" management
    • An isolated, pre-timed location. Not much capability to adjust to traffic demand.
    • Problems: no way to clear congested queues; not responsive to upstream demand
  • Local Control 1970-80
    • Fixed segments of 'upstream+ramp+downstream' sections of highway using detectors to verify success
    • Problems: not responsive to downstream bottlenecks that would back up
      A "Local (ramp) Control" graphic is depicted as a linear freeway, using three detectors; one upstream of the on-ramp, one on the ramp, and one downstream of the ramp.
  • Coordinated 1980-1990
    • Improvements on local control; use of TMC's; greater sophistication
    • First use of algorithms (beyond just "timing patterns")
  • Responsive (i.e., "adaptive") 2000's
    • At the most-congested MPO's; can understand multiple and dynamic bottlenecks.
    • Uses real-time data in 30-sec or 5-min intervals to readjust the algorithms
  • Predictive Future?
    • In theory, would use upstream changes in traffic density to predict conditions and "forewarn" the meters how to operate

Slide 4

Where are R-Meters used today?

Representative (not all-inclusive) as of 2012

  • Most Robust: (i.e., have the most installations, largest deployment)
    • So. Cal; NY-NJ; Chicago; San Fran; Minn; GA; San Diego; Seattle
  • Others: (i.e., mid-sized cities and/or moderate # of ramps)
    • Miami; Phil-NJ; Houston; Phoenix; Portland;
  • Small metro areas: (i.e., smaller regions or small # of ramps)
    • Cincinnati; Kansas City; Las Vegas; Columbus, OH; Salt Lake City; Denver
  • Entire U.S.:
    • 28 of 101 Metropolitan Regions
      • 12 of 15 "very large" . . generally 3M population or greater
      • 11 of 32 "large". . generally 1M to 3M in size
      • 3 of 33 "medium" sized . . generally 500K to 1M population (Baton Rouge, Allentown, Fresno)
      • 2 of 21 "small" sized . . generally 150K to 500K population (Madison WI, Provo UT)

Slide 5

Evolution of Ramp Metering

An X-Y graph titled "Evolution of Ramp Metering", which summarizes that the use and complexity of ramp metering has increased over seven decades.


Slide 6

Push Backs and Challenges

  • Ramp meters were removed or deactivated after being installed in Dallas, San Antonio, and Austin, TX
  • Other cities (e.g., St. Louis and Phil, et al) have removed some, kept others
  • In 2000, MN legislature mandated a "recall" of use of RM's, resulting in a $650K study.

Slide 7

Push Backs and Challenges

  • RM's do a poor job in inclement weather and during special events
  • Queue back-ups force "clears" or overrides that effectively restart the algorithms
  • Challenges exist in properly staffing, training, and implementing RM's
  • Public acceptance is still an issue
  • Agencies and peers have done a poor job of marketing the benefits and relatively high return on low investment

Slide 8

Picture of a red signal indication with adjacent sign "One Car Per Green".
Minnesota Ramp Meter Study — 2001

Results of 2001 study of Ramp Metering Effectiveness

In September 2000, all 430 ramp meters were turned off in the Twin Cities region in response to a mandate from the MN State Legislature, following citizen complaints and questions raised by State Senator Dick Day; namely, do ramp meters work?

Objectives

  • To fully explore effectiveness of ramp meters; meter "ait time" was also a key concern
  • To respond to citizen's questions and identify public perception of ramp metering
  • To involve a citizens advisory board to ensure credibility of the study

Process and Findings

Cambridge Systematics was hired by MnDOT to perform the 3-month study, inclusive of getting pre-study data and incorporating any/all citizen input and ensuring a transparent process. Five weeks of "before" speed and crash data, et al, was recorded. The ramps were shut off for a pre-determined "transition" period and then turned back on for five weeks of "after" data gathering.

  • Without meters
    • A 9% reduction in freeway volume; a 22 % increase in travel times; a 26% increase in crashes (even after adjusting for prior seasonal rates)
    • Most survey respondents believed traffic had worsened with meters off
    • After the study: 20% wanted meters left off; 10% want them "returned"; 70% want modifications

Lessons Learned / Changes Implemented

  • Neither "all" nor "nothing" was deemed best, but a new, modified approach was adopted:
  • Fewer meters than before the study were turned back on (location candidacy was tightened and superfluous meters were removed)
  • Hereafter, meters would wait no more than 4 minutes on local ramps or 2 minutes on freeway-to-freeway ramps
  • Vehicles queued back to city streets will be "released" (meters temporarily shut off) and meter operation will better-respond to congestion-only times via improved use of detectors

Slide 9

Ramp Metering Benefits

Location Safety Congestion Mitigation
Portland, Oregon 43% Reduction in peak period collisions 17% in average travel speed
Minneapolis, Minnesota 24% reduction in peak period collisions 16% in avg. travel speed; 25% increase in peak period volume
Seattle, Washington 39% reduction in collision rate 52% increase in avg. travel time; 74% increase in volume
Denver, Colorado 50% reduction in rear-end and side swipe collisions A 57% increase in average peak period travel speed and a 37% decrease in average travel time.
Detroit, Michigan 50% reduction in total collisions; 71% reductions in injuries An 8% increase in average travel speed and a 14% increase in traffic volume.
Long Island, New York 15% reduction in collision rate A 9% increase in average travel speed

Source: FHWA Ramp Management and Control Handbook, 2006


Slide 10

Recommendations for start-ups

  • Agencies should start small (one or a few ramps)
  • Conduct pre-analysis to
    • ensure candidate locations and deployment exists
    • gather "before" data to compare to "after"
  • Instill public acceptance
  • Become "ramp meter smart" via training and experience before expanding the system
  • Make sure a strong deployment of detectors exists or will evolve
Office of Operations