KiE Square Consulting addresses the Delhi-Gurgaon Toll Gate Imbroglio

Sept. 28, 2012 - PRLog -- Advanced analytics company KiE Square Consulting has come up with an innovative application of queuing theory principles to address the typical traffic-jam situations at the Toll collection Plazas. With the Gurgaon-Delhi expressway management coming under a lot of flak from the government as well as commuters for not applying enough mechanisms to control the errant traffic conditions, there have been a bundle of proposed e-way reforms from several quarters.

Dr. Kamaljit Anand, Co-Founder, KiE Square Consulting and respected Fellow of Indian Institute of Management, Ahmedabad, shared that while promoting electronic tags adoption is a reasonable long term idea, structural proposals like widening of the toll plaza is an expensive and non-scalable proposition especially when an effect of similar nature can be easily created with minimal interventions in the current system. Having experimented widely with typical bottleneck situations on manufacturing shop floors, railway tracks, broadways and metro stations across contexts, he brought out the fact that in most similar situations the most intuitive solutions get deployed even though they may not be the most rewarding. He explained that a Toll gate system's efficiency is measured through two key indicators queue length and the waiting time. The queue build up is a result of the gap in vehicle clearance vis-à-vis total arrivals at the toll per unit of time while the waiting time is a manifestation of the queue length and the time of arrival into the queue. These parameters swell exponentially with the delay in joining the queue in a peak hour.

Ayush Biyani, business consultant at KiE Square Consulting indicated that during Peak Hours there is a continuous queue build up at the rate of around 2.6 vehicles per minute per toll lane and it continues till the arrival rate decelerates. Arrival at various points in the peak hour means a varying queue length and hence a variable waiting time ranging from 5 minutes at the onset of the peak hour to about 26 minutes after one hour of peak traffic with as many as 200 vehicles in queue per lane. One non-working toll booth adds about 7 vehicles per minute across the queues of other toll lanes and in 1 hour of peak traffic, it increases the waiting time at other toll gate lanes by about 5 minutes each. This delay gets proportionately multifold with the number of non-functional gates. The non-peak hour queue build is negligible even with 2 non-operational gates. It is therefore not necessary to keep all gates operational during non-peak hours to optimize resources costs to manage the toll.

Proposed Improvement

Dr. Anand explained that the Current Toll gates have only one clearance feasibility at a point in time, so a maximum of 16 vehicles can be in the process at a time. If the same could be increased to 2, 3 or 4 booths per lane without widening the toll plaza, there can be significant impact on the queue lengths as in effect they have a potential to emulate a 32, 48 or 64 lane toll plaza, albeit with some inefficient allocations of vehicles across the booths within a lane reducing the actual effectiveness to some extent. Continuous widening of the toll plaza road (Single booth Parallel Gates) has its physical limitations so we experimented with the incorporation of additional booths at calculated distances in a single toll gate lane (Multiple Booth Parallel Gates) in a simulated environment. The introduction of multiple booths allows for multiple clearances in a single cycle as against single clearances in the current scenario, leading to a faster clearance rate per minute and a lower queue build up.

Upon initiation of the exercise, we hypothesized that it may take up to 4 booths in a lane to reduce the queue length to a manageable number. However, the eventual findings were extremely lucid in indicating that no more than 2 booths are required per lane to reduce the queue length to near zero within 30 minutes of peak hour, which is a significant achievement if implemented. Two Booth parallel gates design completely nullifies the current wait time of 26 minutes seen at one hour of peak rush and hence can play a significant lever in managing the toll traffic. As the number of booths per lane is less, the inefficient vehicle allocation problem is also minimized to a large extent. The operator can flexibly decide on the number of toll booths to be kept operational during peak or non-peak hour, given the arrival rate but is advised to do it only with the help of an automated decision support system. The effort should also focus on minimizing any non-operational Toll gate lanes during peak hours as it rapidly builds the queues.

Other industry experts and operations researchers have welcomed the approach suggested by KiE Square Consulting and encourage the government to experiment with the same before zeroing on to any large scale mechanisms to address the problem that involve operational challenge, and a large time and financial budget commitment in time of economic pressure.

According to Dr. Anand, the number of lanes in the toll plaza shall not be increased and the planner shall improvise the current system through inclusion of one additional toll booth in each toll gate lane to completely mitigate the queue build ups. There has to be every attempt to save as many person minutes as possible and stop the cascading effect of the highway traffic on to the service lanes, arterial roads and alternate routes. The first, foremost and the only step towards that is a scientific traffic handling system. It is only a fair ask to facilitate the commuters and the investors with it, as they make a choice in Gurgaon amongst several other promising satellite towns around the national capital.

The detailed findings are shared in Economic Times, 27th September 2012