The SMS had piqued M Ramesh’s curiosity.
On a Wednesday night in October 2016, the Divisional Controller (East) of Bengaluru’s bus service had received a complaint about a missing bus.
A commuter, who did not leave his name, complained that a bus scheduled to leave Hoskote at 8 pm was missing. He had waited for over an hour at his stop near Hoskote to take him to Domlur, a neighbourhood near the city centre.
From the tone of the message, Ramesh inferred that the commuter was upset. Hoskote lies 25 kilometres from the city centre, and at that time of the night, the journey back would’ve been trying.
Ramesh forwarded the complaint to the depot manager supervising this route. He reported that everything went as per schedule on the day of the complaint. The crew had recorded sales for Hoskote to Domlur route and had even deposited Rs 1,300 from ticket sales. They had returned with the bus to Hoskote at 3 am the next day after halting for the night at Domlur.
In normal circumstances, Ramesh would have let it go. But he had a hunch that there was something else going on. He decided to dig deeper.
A few months prior to the complaint, the Bangalore Metropolitan Transport Corporation (BMTC) had started implementing something it called the Intelligent Transport System (ITS) project to modernise the city bus service. Among several other things, this enabled them to record, using GPS, the location of buses every 10 seconds. The ITS also logged where the bus stopped at what time and for how long.
So Ramesh logged into the ITS server and pulled up the said Hoskote-Domlur trip from the server. On a web browser, a map of Bengaluru was overlaid with two routes. A zigzag green line indicated the Hoskote-Domlur route that the bus should have taken. Below it, a zigzag red line showed the actual route it took.
Ramesh zoomed in onto the map. Though the bus had started near Hoskote, it drove to Benson Town, about 10 kilometres away from Domlur. The bus had also halted at a community hall in Benson town that night, instead of Domlur as claimed.
The crew had lied. Ramesh guessed that they had made an unauthorised trip for someone. Charter trips are not uncommon but this one was off the books. He checked the routes that bus had been taking for every trip over the entire week and found out that their deviation was a one-off incident.
BMTC confronted the driver and conductor, who admitted that they ran a charter service for a private firm in Benson Town. The company paid them Rs 8,000 for making the trip. To cover their tracks, they deposited Rs 1,300 at the depot for the shift that BMTC had assigned to them.
BMTC suspended the driver and conductor for a few days. The mystery of the missing bus was solved.
A small auditorium at BMTC’s central office at Shanthinagar bus station in central Bengaluru is bustling with activity and phone calls.
There are 20 seats with a computer and a phone on the desk. At the head of the room is a video wall — six 55-inch LCD screens put together on the wall. Graphs, maps, and tables appear and disappear on the screen.
This is the ITS command and control centre. Staff members in this room track all BMTC buses across depots. For BMTC to automate its operations, it needs to process a massive volume of data.
Each VTU sends 276 bytes of data every 10 seconds as it updates its position. And each ETM uploads 230 bytes of data on ticket sales every five minutes.
So the entire fleet of 6,440 buses, on average, sends 15 GB of data in a day. In a year, the ITS will approximately generate more than 5,000 GB of data. All this data goes to the Karnataka government’s State Data Center at Vikasa Soudha, the state government secretariat. It also retains all historical data generated by the ITS.
Harsha Ramachandra Kanak is sitting in a corner and staring at his laptop. An employee of PricewaterhouseCoopers (PwC), he has been working for the last two years with BMTC as a consultant.
He opens a map on his laptop screen: it’s a map of Bengaluru with many tiny buses on it.
He goes to the vehicle tracking dashboard. It’s a graph with blue, green, purple, red and yellow bars, which are in fact alerts. These show buses that have skipped stops, deviated from their routes, departed late, broke the speed limit, or stopped at unauthorised spots.
He clicks on the red bar, which informs him that 449 vehicles are stationary across the city. A table opens with all 449 buses that have been stationary for over 15 minutes, with their details.
He picks one of the stationary buses and calls its driver.
“Is there a problem? Has your bus broken down?”
“Yes sir, there seems to be a small mechanical problem.”
“Have you called the depot?”
“Yes, the depot is sending somebody to fix it. They should be here in 10 minutes or so.”
Harsha hangs up and calls the driver of another stationary bus.
“Is there a problem? Has your bus broken down?”
“No, sir, there is no problem. I was just going to move.” The screen shows the bus has started to move even before Harsha has hung up. The VTU on the bus is doing its job.
There is another set of data coming in from the ETMs. Harsha clicks on bus number V-335E. The table that rolls open shows it is on its second trip for the day and has collected Rs 2,681 by issuing 50 tickets so far. With another click, a table rolls out with the number of tickets issued at each bus stop.
The origin of the ITS goes back to September 2012 when Anjum Parwez, an Indian Administrative Service officer, took over as the Managing Director of BMTC. A graduate in electrical engineering from IIT-Delhi, Parwez had developed an early interest in computers and information technology.
At his first IAS posting in 1994, he got a few students to develop a website for his department. This was just three years after Tim Berners-Lee had launched the first website.
Later, as the CEO of Belgaum Zilla panchayat in, he got the entire inventory of the district administration computerised. “My hobby is software. While trying to understand the operations at BMTC, what struck me was the lack of use of IT in urban transport sector,” he says.
By 2012-13, the Union government began looking at funding pilot projects in ITS projects. Some cities experimented with IT solutions in a fragmented manner with varying success — Indore with ETS, Ahmedabad with smartcards, and Bhubaneshwar with fuel monitoring.
But Parwez, with the largest and most profitable fleet in India, wanted to try something more ambitious. First, he created the post of Director (IT), and Indian Forest Service officer Kumar Pushkar, his batchmate, was appointed.
“Pushkar and I started brainstorming. Our ambition was to harness the potential of IT to benefit internal operations as well as commuter services,” says Parwez.
But, some on the BMTC’s board of directors were not convinced. “Our job is to operate buses, not give some world-class technology service,” Parwez recollects one board member responding.
“BMTC has been the only profitable public bus service in the country. Why should we change anything?”
“This has not been tried at this scale anywhere in the world.”
“The employees will protest.”
“A change is a big risk.”
“Technology involves huge costs; it’s a loss-making venture.”
The objections came fast and thick, but Parwez and Pushkar stuck to their guns.
“My simple answer to them was it might not bring in immediate revenue enhancements but the spin-offs are multiple. ITS would help prevention of pilferages, monitoring of vehicles, optimisation of fuel consumption, and understanding the behavioural pattern of the crew,” he says.
The board reluctantly agreed. The team started parallel work on all fronts, but first, BMTC needed a technology firm to execute the project. Many expressed interest and promised to bid.
In 2013, the BMTC issued a request for proposal (RFP). It was taken aback when it received hardly any bidders. It settled on a firm, which quit the project within a month, even if it meant losing their initial deposit of Rs 50 lakh. The company had not fully understood what the project would entail and had trouble finding customised solutions for BMTC.
“No IT project of this scale existed in the transport sector and no firm had known or readymade solutions. Another deterrent might have been that the RFP included requirement for flexibility (to alter specifications/ requirements at a later stage) as we were still figuring things out," says Parwez.
BMTC called for a re-tender in soon after and Trimax IT Infrastructure and Services, a Bengaluru based small firm, was selected.
Trimax provides 6,500 VTUs, 10,000 ETMs, a data centre, a control room, and around 100 employees to work on ITS. Under this PPP project, BMTC does not own any of the equipment and stays free of the liability to maintain the system. Between 2016 and 2021, it will spend Rs 70 crore to build this system on a Build-Own-Operate-Transfer model. On completion of the five-year term, the entire project – all the equipment, all the data, the IP, and the application – will be owned by BMTC.
In the initial days, challenges were many. A project closer home in Mysuru provided some valuable insights to the BMTC. Mysuru's ITS received funding from the World Bank and technical assistance from tech giant Tata Consultancy Services.
It was a small project with generous funding and sound technical support. One learning was the need for a rugged industrial GPS to work on bumpy bus rides.
But, the Mysore project was a small one involving 500 buses and 105 bus stops. Whereas BMTC’s network was more than a dozen times bigger — 6,440 buses, 7,000 bus stops, 43 depots and a workforce of 36,000. No Indian firm had ever worked on an ITS project of this scale.
One of the first tasks was to map each bus stop. BMTC began assigning GPS co-ordinates against the name of the 7,000 bus stops. But when it used this data on Google Maps, it threw up an odd mismatch. Bus stops which were 1 km apart on the ground were shown 18 kms apart on Google Maps.
Bengaluru had many bus stops with the same name and this confused the system. To get over this, BMTC used codes for each bus stop instead of names.
Once it fixed this, bus corporation hit upon another hurdle. Often, the map was showing longer roundabout routes, when shorter routes existed.
“We realised we were not getting dynamic maps from Google. They keep updating the map but don’t put it on the open system,” says Parwez. BMTC paid up to get access to Google’s dynamic map.
In October 2013, BMTC hired PwC as a project management consultant. By August 2014, it saw to it that VTUs had been installed on some buses. It finalised ETM specifications. It automated backend operations at the depots and installed computers in the central control room. It even came up with a name for the mobile app.
That is when Parwez and Pushkar were transferred out of the BMTC. The future of ITS was up in the air.
To replace Parwez and Pushkar, Ekroop Caur and Biswajit Mishra came in as MD and Director (IT). In 2015, BMTC conducted a survey on commuter’s expectations for public transport. The survey results reaffirmed that the direction ITS was taking was the right one.
So the management doubled down on completing the project.
Once the ITS was ready to use, the most critical challenge was not technical. Like in projects of this scale, the challenge was a behavioural one.
Drivers, conductors, and depot managers viewed the ITS with suspicion, as a system to track and penalise them, says Mishra. In the initial days, the ITS team found many VTUs damaged, either broken or burnt. Some were even found clogged with water.
Mishra recalls one incident when he noticed that the ITS was showing a bus in the Middle East. He assumed there was some problem with the GPS device and it would need to be replaced.
“The guy who went to replace it gave me a call from there. He said: ‘The GPS device is fine, but these guys have put a metal box on top of the VTU.’”
The crew figured that if they covered the device with metal, it would give them an erroneous reading and the ITS would not be able to track them anymore.
“Look at the ingenuity!” he said. “So, we knew well that we needed to involve them; taking strict action was not an option.”
Conductors did not take to the new ETMs well either. The ETMs would lose all data if dropped. They also ran out of charge frequently. And, then, there were conspiracy theories. “Like it was a China-made machine and therefore, unreliable or even harmful,” Mishra recalls.
Caur and Mishra held a series of meetings with the staff. They explained that the aim was not to keep a watch on them but to improve operational efficiency. They installed charging units for ETMs and told them that there’s nothing wrong with using a Chinese machine. “I told them many smartphones used today are made in China,” says Mishra.
Slowly, BMTC conductors, drivers, and other staff came around.
On the technical side, the BMTC faced many challenges in building the ITS.
For instance, when over a hundred buses were at a depot, the GPS would wander due to interference from the metal bodies of the buses. BMTC worked around this by increasing the tolerance for error in GPS readings. “We did a lot of such jugaad to ensure it worked well for us,” says Mishra.
BMTC also had a tough time getting the ETMs custom made for them. They needed a combination of a ticketing machine and a Point of Sale (PoS) machine, like the ones used in supermarkets. The device had to be portable, rugged, and equipped with GPRS.
None of the existing machines on the market were suitable. Together with Trimax, Verifone, one of the top manufacturer of PoS terminals, re-engineered their PoS device to meet BMTC’s needs. The ETM used by BMTC today was developed after ten revisions.
Designing the smart card was another tricky issue. This is the first time in India a transport agency is using an open-loop smart card. Closed loop smartcards such as ones issued by Bengaluru’s Namma Metro can only be used on the rail metro service, while open-loop smartcards can be used by multiple agencies, including as a debit card. The Reserve Bank of India had even come out with guidelines for smart cards for mass transit systems. Each part was tested, verified and then notified by the RBI. This took a lot of time.
But BMTC’s biggest challenge was in designing around internet connectivity issues. In the web-based model, syncing thousands of VTUs and ETMs needed 99% uptime while connection would be down for as long as 24 hours at times.
So BMTC went back and re-designed the ITS. The new version requires internet only when the depots need to sync data with the server. This syncing of data is done just once a day, usually at night, when connectivity is good.
With the ITS, BMTC is making a significant commitment to base decisions, both big and small, on data.
With bus tracking, BMTC has begun preparing bus schedules that make sense in today’s traffic. The bus schedule in use was prepared in 2011, five years before ITS. And it does not take into account traffic growth or traffic variation through the day.
For instance, the bus schedule requires a crew to make six trips in a day for Route 335E between Majestic and Kadugodi. This assumed that travel time between the two places was 90 minutes. But in reality, travel time has increased to two hours since then.
So, either the crew fails to complete their assigned trips or have to put in unpaid overtime. But with ITS, BMTC can record the time taken for each trip and allot time, based on this data. Moreover, they can even assign time taken for the trip based on the time of day.
Today, BMTC allots Volvo buses on Route 335E 80 minutes during lean hours and 2 hours during peak traffic hours. From six trips in the original schedule, now, the crew needs to make only four complete and two partial trips.
BMTC is doing this analysis for all routes to rationalise schedules.
It is also working with IIM-Bangalore to rationalise routes based on demand in an area and remove redundancy (multiple buses on the same road without enough demand). This could simplify the existing complex network of over 2,400 routes and encourage new riders to take the bus.
Another key goal of ITS has been to improve the performance of its employees. Automated reports generated by ITS has played a big part in the uptick in productivity at BMTC.
Since April, the central office has been calling in each depot manager, and giving them a report of the worst performers. “We have decided to focus on one parameter every month. We made it clear to them the worst schedules have to be handled first and subsequently the discipline would percolate,” says Mishra.
These automated reports made redundant five divisional offices that operated between the depot and the central office. Now, those 400 employees work at the central office.
BMTC is also focussing on real-time optimisation of their services. And this is where the control room comes into play.
“You notice that a bus is overspeeding, stationary or deviating from its route. Immediately pick up the phone and ask the driver,” says Mishra to his control staff. “You can correct the non-adherence right then. You do not have to wait for the monthly report to come to take disciplinary action.”
Mishra wants the control room to be like a “machchli bazar (fish market), bubbling with activity. He wants each operator in the 20-seater control room to track activities of 2-3 depots. For this, the BMTC is building a two-way communication radio for its buses. It is installed to the right of the driver’s seat so that he can access it easily.
One use of the radio is to prevent bunching of buses at a stop. The control room can ask buses to move fast or slow down to space them better and reduce waiting time to 2-3 minutes.
Bus drivers can also use the radio in case of a medical emergency or a mechanical breakdown. “Before twitter handles take pictures and abuse BMTC, the control room can clear the broken down bus,” Mishra says, smiling.
Data also plays a major role in tweaking fares.
ETM data showed the number of “second stage trips” (distance of two to four kilometers from the origin) had fallen since April 2014 when fare was hiked from Rs 10 to Rs 12. Such trips account for over 25% of BMTC’s trips.
In April this year, BMTC decided to drop the fare back to Rs 10. This could’ve implied a loss of around Rs 25 lakh. But after the revision, the revenue shot up as the number of second stage trips went up by one lakh.
This counter-intuitive outcome was not down to luck but one based on analysis of data over six months. And, it was not a tentative move, it was a confident decision. In the original fare revision, when BMTC raised fares for all trip distances only second stage trips had shown. It was strange. The officials realised it was a case of revenue pilferage. When passengers didn’t have change, the conductor would accept a 10 rupee note and not issue a ticket.
Once the bus utility reduced the fare back to Rs 10, payment became easier, trips were registered, and the pilferage ceased.
Based on ETM data, the BMTC is also going to introduce Happy Hours — reduced fares during off-peak hours — on routes with low ridership.
The ITS has also had benefits in surprising areas. Take accident claims for instance.
Mishra recounts receiving a call about an accident involving a BMTC bus recently. The complainant alleged an overspeeding BMTC bus had hit his car.
Over the last few years, BMTC has had to pay compensation running into crores in hundreds of accident claim cases. In this case, the BMTC driver was insistent that he was not involved in an accident.
So Mishra decided to check in with the control room.
The data from VTU showed that at the time of accident, the bus was not close enough to the accident spot to have hit the car. The bus wasn’t speeding either.
BMTC shared the data with both the complainant and the traffic police. The case was closed.
“Blaming the BMTC for an accident makes it easier for traffic police to sort out the case, and for the complainant to receive compensation,” says Mishra. “ITS data makes it very easy to confirm the genuineness of claims.”
With ITS, BMTC’s accident liability has come down by 23% between 2014 and 2016. While BMTC paid out for 388 accidents in 2014, in 2016, it paid for only 299 accidents. It had paid out Rs 11.5 crore in accident compensation in 2014.
The BMTC smartphone apps on both iOS and Android are already reaping some of the advantages of ITS. And for commuters, the most prominent feature is the arrival time, or ETA, using location data from the bus.
To get accurate ETA, BMTC had to go through multiple iterations.
The first version calculated ETA using an in-house algorithm using VTU data and travel times that they had collected. But, this produced ETA inaccuracies between 20 minutes and one hour. Another version with a team from IIT-Madras did not do much better.
Finally, BMTC passed on real-time GPS coordinates to Google Maps and let it calculate the ETA. Google Maps API uses a combination of real-time data and predictive analysis based on its own historical data. The accuracy was quite high — the error in ETA was within 2 to 3 minutes.
BMTC is also displaying ETAs on electronic information boards at bus stands and bus stops, similar to the boards at railway stations and airports. These Passenger Information Systems (PIS) will let you know when your bus is going to arrive at your stop.
Though useful, the PIS has turned out to be the weak link in the ITS. Only 35 PIS have been installed at 11 locations including major bus stations and airport. The lack of 24/7 display boards, network connectivity downtime, and lack of funding have limited their spread.
Another hindrance is a turf-war between two city agencies. BMTC owns only 55 of the total 7,000 bus stops across the city. The rest are owned by the Bruhat Bengaluru Mahanagara Palike (BBMP), the city municipal corporation. And, the two agencies have been squabbling over who should put up the PIS.
The BMTC has recently invited a tender for PIS at 400 locations including metro stations. “Once we succeed in running PIS at these 400 locations, BBMP would be under public pressure to install PIS across the city,” says Mishra.
Another vital component of making bus transport seamless is cashless transactions. By July, when Parwez came back briefly as MD, BMTC had begun a smart card trial on an A/C bus route that services the IT corridor. Many enthusiastic commuters volunteered to try the smart cards and provided feedback to tweak and fix some issues. BMTC also created a WhatsApp group with officials and passengers to fix problems.
Parwez recounts a commuter’s text informing that the conductor issued nine tickets instead of one when he mistakenly pressed ‘9’ instead of ‘#’. “We redid the software. It prints a ticket only on pressing the # key.”
BMTC’s open-looped smartcards can be redesigned to enable you to commute on the metro, inter-city bus services, private taxis, and autorickshaws. This makes the BMTC ready for a future of integrated multi-modal transport.
And in the future, the ETMs could receive payments from mobile phones using NFC connectivity. “I hope, someday, BMTC will also allow you to pay by Paytm or another wallet,” says Parwez. He was replaced by V Ponnuraj by the end of July as the MD.