CITMS: A mathematical solution to the traffic management crisis

In 2002, Md Ashraful Alum was a fourth-year student in the Department of Architecture at Bangladesh University of Engineering and Technology (BUET). He resided in Dr M A Rashid Hall, with his roommate Shuvo, a junior from the same department. 

Despite the age gap, they had a friendly relationship.

One day, Ashraful received a call from Shuvo's number, but someone else was on the line. The voice said, "He has died in a road accident. You may identify him since you were the last person he spoke to. Please come to the morgue."

A courier van ran over Shuvo in the Jatrabari area of the capital.

This incident left a profound mark on Ashraful. He promised to himself that he would work on road accidents for at least a year. He would conduct a root cause analysis of the issue. 

Ashraful lived up to it. He took to the streets. He began observing the movement and behaviour of vehicles on the roads. In this way, about two and a half years passed. He witnessed firsthand the full scope of the capital's deadly traffic system.

Issues such as mechanical defects in vehicles, unskilled drivers and frequent law violations came to the fore.

He identified some typical reasons behind road congestion between two signals. These include setting up shops or parking vehicles on the road, stopping vehicles in the middle of the street, pedestrians crossing the road carelessly and the competitive attitude of public transport drivers.

"I observed that the drivers didn't have any feelings about the accidents or congestion. It's a systemic failure of our society," Ashraful told The Business Standard.

"To understand the importance of solving a problem, we first need to realise how much harm it is causing. I believe that traffic jams and road accidents are silent killers of our economy. But since we don't directly see the damage they cause, we fail to feel it. 

As a result, we learn our lessons at the cost of tremendous losses," he said.

We need to view the traffic problem from a mathematical angle. It must be solved using numbers and data. Our system has huge business potential. If we charge only Tk30 monthly in return for complete traffic alerts, I believe around 10 lakhs of people will pursue it.

Although he did not come up with any solutions at the time, Ashraful initiated some awareness-raising activities, such as organising photography exhibitions, providing advanced training for drivers, etc.

Afterwards, he became preoccupied with his career as an architect. Although he was driven by a sense of responsibility to work on this issue, he took a long break, disconnecting from road safety initiatives. Instead, he focused on his career. 

However, something notable happened in the meantime. Ashraful identified 20 blackspots along the Savar-Manikganj highways and shared his report with the Superintending Engineer of the Roads and Highways Department. Just one month later, the infamous accident that killed Mishuk Munir and Tarek Masud occurred at one of those spots.

The year 2017-18 marked a turning point. 

He delved back into road safety — or perhaps the matter found its way back to him. He learned about a project by the Asian Development Bank (ADB). He proposed a modernised signal system at that time.

He continued to work on the issue again.

Fast forward to recent years, Ashraful has made a team and come up with an integrated Artificial Intelligence-based solution that they call CITMS, short for Customisable Integrated Traffic Management System, through his company TrafFix.

How CITMS works

The core idea is to keep all informed about traffic situations so that they can cooperate seamlessly. That's where data comes into play.

"We want to involve the public in our system. Anyone can help by providing traffic information using our system. If there is severe congestion or any disruption on the road, they can capture a photo or video and send it to our system. Our team will then try to reach the scene to address the issue," Ashraful said.

There are three major parts of the CITMS solution — signal, camera and operations.

The signal system will be both manual and automated. If any ambulance or VIP movement happens they can operate manually. Rest of the time, it will be fully automated with condition-based timing.

Each signal will have cameras providing a real-time view. Traffic police officers can input information into the system from their respective locations. If a road is closed due to an accident or repair purposes, everyone will be informed beforehand so that they can avoid the route. 

It can also serve as a communication medium among traffic police officers, replacing walkie-talkies. If the power system or internet is interrupted still it will run as it has four levels of backup system. So, even blackout or internet shutdown will not affect signal operations.

"The system can record activities and efficiency of traffic police officers. Thus it will make them more accountable," Ashraful said.

In 2023 alone, 7,902 people lost their lives, according to data from the Bangladesh Jatri Kallyan Samiti. A survey by the Centre for Policy Dialogue (CPD) last year revealed that a Dhaka city resident wastes about 276 hours annually sitting in traffic congestion.

"We need to view the traffic problem from a mathematical angle. It must be solved using numbers and data; otherwise, we won't be able to find any solution," Ashraful said.

TrafFix consists of 16 people, including four founding members– Dr Mohammad Nurul Hasan, Abu Anas, Mahedi Hasan, and Md Ashraful Alum.

Most of the members are skilled software engineers.

Ashraful mentioned that three to four types of experiments have been initiated by various govt organisations, investing hundreds of crores but all of them failed.

However, CITMS, he claims, is future proof — meaning it's designed like 'open architecture' to fulfil future requirement demands. That's how the team has been relentlessly working to upgrade the system over the past few years.

As the word 'integrated' in the name suggests, the centre point of the solution is a mobile application. The app features different modules to bring together various stakeholders, including the city corporation, police, BRTA, journalists and government agencies.

Artificial intelligence will handle tasks such as assessment of traffic flow, monitoring the average speed of vehicles, understanding the effectiveness of signal timing, automated data analysis of traffic signal efficiency and identification of vehicle number plates. 

Additionally, it will handle tasks such as the detection and case filing for traffic signal violations, illegal parking, illegal lane change and speeding, and data analysis of traffic congestion, etc.

"Our system has huge business potential. If we charge only Tk30 monthly in return for complete traffic alerts, I believe around 10 lakhs of people will pursue it."

The solution comes at an affordable cost as the raw materials are locally sourced. Though it may vary, the device, lights, and accessories will roughly cost Tk25 lakh to Tk50 lakh — nearly one-tenth the price of imported ones.

Ashraful believes that we can even export the system to other countries. If the government invests in this, it will generate returns in less than two years.

"We have shared our project with experts and a few stakeholders, and they see it as a robust and sustainable solution. However, we are currently awaiting discussions with officials of the new interim government," Ashraf concluded.

TrafFix wants to work with the government for two years to implement its solution and strengthen the government's capacity. After that, the government will take control and keep them as shadow consultants.