ATCC 96.5% accurate on Indian roads

ATCC 96.5% accurate on Indian roads

An independent study has shown that MetroCount automatic traffic counters and classifiers (ATCC) provide average accuracy of 96.5% when monitoring vehicles on Indian roads.

This incredible result is a testament to the 3 decades of tireless research, development and continual improvement of systems by MetroCount engineers.

ATCC 96.5% accurate on Indian roads

Speed accuracy on Indian roads

Not only was the average accuracy for volume and class of vehicles recorded at 96.5%, but speed accuracy was also measured at over 95%.

The independent researchers reveal this is “exceptionally good considering the Indian traffic conditions” which is plagued by inconsistent lane discipline.

Why ATCC are more practical than video counts

This research was conducted in response to many drawbacks of video traffic monitoring in India. Researchers wanted to find a more reliable and practical method for counting, classifying and monitoring the speed of vehicles on the roads. Reasons for this include:

Drawback of video counts

  • Video counts require expensive camera equipment to be set up in public needing security protection or a safe locking mechanism to avoid theft or vandalism.
  • Cameras are limited to recording for short periods due to storage and power requirements. This allows for a very small sample size of data collection.
  • Raw video is watched back at a later date by manual counters. This allows for the possibility of human error and fatigue. It can also increase the final cost of the data collection process.
  • The camera angle may not allow for clear viewing of all lanes of traffic at all times.
ATCC 96.5% accurate on Indian roads

Considering these drawbacks, researchers concluded that automatic traffic counters were an exciting possibility for replacing manual and video counts to monitor traffic in India. Reasons for this include:

How ATCC solve these issues

  • Automatic traffic counters are small, discreet systems that can be set up at the roadside and securely left for weeks without supervision. This is thanks to two separate protective layers and locking mechanism ensuring vandalism, water, traffic impact and even fire will not affect the data collection.
  • MetroCount ATCCs can run continuously for up to four years without requiring a battery change and store information on 1 million vehicles before requiring data download.
  • MetroCount MTE® Software does all the time-consuming analysis work, saving resources and ensuring no human error or fatigue is involved. Encrypted algorithms ensure data is secure and unable to be tampered with at all time.
  • Pneumatic tube sensors used with MetroCount ATCC collect data on every single axle that runs over them, timestamping, counting, classifying and monitoring the speed of every vehicle.

Having come to these conclusions, the final step before confidently commencing the use of MetroCount ATCCs was to test their accuracy and reliability. Testing was completed by the Department of Civil Engineering at the RV College of Engineering. Read below for the full research report or click here to see the original report document.

Original report

“Department of Civil Engineering, RV College of Engineering, engaged a performance evaluation test on the product supplied from MetroCount, Australia / 3E Consultants Indian representative with a model name of RoadPod® VT 5900 and RoadPod® VT4 5904.

It is hereby found that the initial performance evaluation results of classified count using automatic traffic counter and classifier equipment are satisfactory.

Similarly, the accuracy of the classified count was checked for the limited categories of vehicles, as MetroCount classifies the vehicle type, direction, axle length and wheel base. However, in Indian traffic conditions the axle length of buses and 2-axle trucks/6-wheeled LCV are almost similar and this classifies as one category for above-mentioned categories of vehicle. Same in case of small cars and 4-wheel LCV/4-wheeler shares autos.

Speed prediction was tested for MetroCount systems obtained from speed measuring systems; the analysis reveals the speed predictions of MetroCount systems is nearly 95% accurate.

Below are the performance observed during the initial evaluation in a location at the Bangalore University Campus, Jnanabharathi, Bengaluru.”

Evaluation results

SL
NO
Vehicle TypeVideo CountMetroCount% Accuracy
12-wheeler3,2813,23098.45%
23-wheeler47745595.39%
3Car1,4271,39797.90%
43/4 wheel LCV11811395.76%
5Bus or truck413995.12%
Total average96.52%

“The performance leads to an overall accuracy of more than 95% when the time gap between each vehicle is more than six seconds. Perhaps in case of time gap less than 3 seconds, accuracy will be 90-95%, which is exceptionally good considering the Indian traffic conditions. The MetroCount and 3E Consultants team of the company gives very good technical support in case of any technical issues.”


Thanks to Anand Kumar B.G M.E, Ph. D, Assistant Professor of the Department of Civil Engineering and Dr. Radha Krishna, Professor and Head of the Department of Civil Engineering at RV College of Engineering for sharing their finding.

Leave a Reply

Your email address will not be published. Required fields are marked *

To improve your browsing experience, this website uses cookies.

Cookies are small text files that enable us to record your website usage in accordance with our privacy policy.
Please confirm if you accept our tracking cookies.