Getting Rid of 2D Chaos — Tackling the Traffic Congestion in Mumbai
Abstract
In this world of connected devices, where each and every movement of ours can be monitored, analyzed and visualised in a matter of seconds, data creation and analysis is supreme. We are creating terabytes of data each and everyday driving on the roads. This data could be used for multiple purposes, one of them being assessing driving behaviour and helping improve it by providing real-time corrections and post journey analysis, an area we are focussing on.
This project aims to change driver behaviour not by creating a fear of authority, but rather through incentivising the driver of the vehicle to drive better, both for his good and for those around him. We are using present day technologies such as proximity sensors, ambient light sensors, cameras utilising machine learning and object recognition, biometric recognition, GPS, etc. to come up with a solution that we hope will be able to induce a change in driver behaviour for the better.
Context
Currently, the traffic on the roads of Mumbai seems to to be getting graver and graver and congestions leading to jams are becoming more of a routine rather than an one-off occurrence. Commuters caught in this charade end up losing a lot of valuable time, wasting precious fuel and hard earned money. Other than these tangible and measurable values, there is also the mental fatigue and stress which has a direct impact on his/her behaviour on the road, which in turn affects the other commuters.
The congestion on roads can sometimes lead to severe implications with regards to emer- gency services. They are unable to make it to their destination which has, at times, fatal consequences for those in need of these services. It becomes essential in such cases to be able to manipulate the traffic to give preference to the Emergency Services.
Thus, it is of vital importance to have a system in place to establish smooth and free owing traffic.
Process
Problem Statement
Given the current situation, we as a group felt something had to be done to curb this madness. We looked towards the internet, went onto the roads, sat alongside drivers, even became the drivers to see what the root cause of this major problem is. After doing this, we came to our first conclusion, ie,
The commuters need to be able to get to their destination in a hassle free manner for a stress free and great experience.
But immediately after we put this statement on the wall, we came to our second conclusion, ie, this statement is too broad to be able define a problem. It became imperative to know what made a journey hassle free, what made it stress free and what led to a great experience.
A hassle free journey was defined as a journey in which the Cognitive load faced by the driver is low, a journey in which they are zero distractions while driving, and a journey in which you are actually moving, not standing still for most part. But more importantly, a journey is hassle free when everyone around you is driving in a predictable manner, following rules and is disciplined. This last point we felt made the difference between nishing the jour- ney with a headache versus coming out of the car refreshed. It became apparent that disciplined driving really helped reduce cognitive friction while driving as everyone’s next step is in some way, predictable.
Stress free and hassle free are correlated. A hassle free journey is stress free — at least for the driving bit of it. Absence of all the things that make a journey full of hassles came to be de ned as a stress free journey.
Great Experience was slightly harder to define. There are loads of small things that come together to make for a great experience. Things like reaching in time, spending less time standing and more time moving, even things like having just the right song playing in the background and seeing more trees enroute to your destination has implications on the experience of the journey.
We took the data that we had and chalked out the needs, challenges and demographics of our Centre-Point User. There was this one extra thing that was included while considering our Centre-Point User — the Driving Habits.
The inclusion of Driving Habits was pivotal in shaping our Problem Statement. It ensured we didn’t look at traffic and congestion just in terms of numbers or red lines, but made us look at the people involved in the whole process. It got us thinking about how we could impact a society at large by changing one person at a time. In short, it got us thinking about Human-Centred Design.
We developed a Persona based on the Centre-Point User (this has been elaborated in the next section). And we changed our Problem Statement to be more user specific.
It read,
How might we help the Commuter Navigate better and Induce Discipline on the road for a Hassle free and Great Experience.
Introduction of the word Discipline all of a sudden got the user into focus. But this was still too blur a problem statement. We were unable to come to a consensus on what our focus area should be. This is when we decided to use the Importance versus Difficulty Matrix to narrow down on a focus area.
This made it clear to us that Driver Discipline was a pressing issue, one that had to be tackled. So, navigation was dropped from our Problem Statement. It now read “How might we help Induce Discipline amongst Commuters for a Safe and Hassle free Journey?”
But this too appeared too broad a Problem Statement. We used the Abstraction Ladder to narrow down on a discipline area that we felt is a major problem in the cities and is some- thing that is solvable. To come to a consensus regarding this, we used the Problem Tree Analysis and found out that most of the indisciplines occur at the traffic junctions and these are easily tackleable. The problem statement now read,
How might we induce discipline amongst drivers at Traffic Junctions to ensure the safety of those around them?
We began thinking of solutions to deal with these problems, when we realised, solving congestion, the topic we initially began with was lost. At the same time we gured that the Problem Statement we had was too narrow. We needed to broaden the statement to come up with a system that can tackle an issue that is a real world problem. And so we climbed up the ladder and got back to our original Problem Statement which became our final problem statement, ie,
How might we help Induce Discipline amongst Commuters for a Safe and Hassle free Journey?
Synthesis of Data
After conducting the primary research and user studies, we started sorting the raw data into different categories. These categories were further divided into sub-categories to segregate the information effectively. Our users were mostly drivers or daily commuters who faced the problem of congestion firsthand everyday. We decided on having 6 major categories with subcategories. These are :-
About the Center Point User : This category dealt with all the information that defined who the user is. It was divided into subcategories such as Demography, Commute Time, Mode of Transport, Attitude, Comfort Level with Technology, Professions and Beliefs).
Their Network : This category dealt with their Family, Friends, Acquaintances, etc who might commute with them from time to time.
Their Needs : This category listed all the needs the user had, both Implicit and Explicit.
Their Driving Habits : This category was divided into two subclasses — their driving habits and their habits while driving.
Their Mental State : This category listed down the mental state of the drivers during and after completing the journey.
Their Challenges : This category was created to add the challenges the user’s faced, either directly stated by the user or deduced from the synthesized and grouped data.
Creating the Persona
While performing user study, we found out certain traits that were common to a lot of people which shaped the foundation of our center point user. Our persona followed these guidelines and was representative of the target audience for our system. Mr. Anshuman Saha is a typical middle aged man who is married and has a 4 year old son. As an office goer in Mumbai, he has to travel to his office in the heart of Mumbai. He has a typical 9 to 5 job for which he has to commute to and from his office during rush hour in his car. His daily commute comes up to around 40km which leaves him fatigued and stressed out after a hectic day at the office. He has recently come to terms with his responsibilities as a family man and wants to become a safer driver while being able to save time and money while having a stress free commute. He occasionally goes out for lunch or dinner with his family in the weekends. If every ‘Anshuman’ on the road uses this system, the chaotic congestion will ease out along with them saving time and money.
Keeping Mr. Anshuman in mind, we went through an initial round of ideation.
Some of these ideas were taken forward into the system we designed.
Final Solution
The system monitors the uses behavior using various sensors in the car. Software analyses the data from the sensors and evaluates the driver’s journey. The evaluation is governed by a point system. For breaking a rule, the driver gets negative points and for maintaining proper discipline on the road, he receives positive points. Each journey is logged with a certain number of points, these add up to a point tally. On the basis of the points, the driver gets a certain discount on the insurance premium.
The product ecosystem also includes a mobile phone application. After every trip the driver is given the journey summary. The summary gives him things he should keep in mind for the next trip he makes, points he earned or any kind of achievements. It also gives a log of all the points lost and gained. The driver also receives an email giving a weekly/monthly report. Details of the whole system is explained in the next section.
System in the Car
The Car comes with company fitted hardware. This includes a few sensors, cameras and a screen on the dashboard. The sensors sense the distance, signs on the road, traffic signals and is aware about its surroundings. On the basis of the data these sensors collect, software decides whether or not road discipline is being maintained by the driver. Accordingly, out of a list of faults, the driver is notified about them whenever he/she is breaking any rule or driving unethically.
The dashboard display is situated in front of the steering wheel. It is placed between the speedometer and the tachometer. (FIG).The dashboard screen gives a visual feedback when- ever the driver makes a mistake. The visuals are designed in way such that the real time alerts given to the driver do not distract him and yet give him the space to correct himself. Visuals are also accompanied by audio feedback in form of beeps.
Ecosystem of the solution
The players -
Insurance Companies : The insurance companies can see the scorecard of the driver and get an overall view of their driving habits. They can then analyze this scorecard and gauge the risk of the driver causing an accident and decide what premium to set for the insurance for that driver. Since a good driver has a lower chance of causing accidents, his premium can be lowered down and this will give him an incentive to continue to drive well. This may result in a significant reduction in the claims, hence saving the insurance company money. It will also allow the companies to identify bad drivers who have a higher chance of causing an accident, and possibly charge them a higher premium on their insurance.
The government : The government is expected to maintain good roads and infrastructure such as proper markings on the roads for lanes, zebra crossings, clearly visible road signs and traffic signals which is essential for the sensors in the car to perform correctly and collect accurate data. In return, the government will see a decrease in accidents generally caused due to bad driving habits ensuring the safety and welfare of its citizens.
Sensors and Point System
The App requires a new driver to register with the system to start gaining points. The new driver gains double points when driving correctly while losing only half if they make mistakes for the rst week. This period of one week is for both the driver and the system to get accustomed to each other. The driver will become habituated to driving with the monitoring while the system learns the areas of improvement for the driver and adjusts accordingly. After the training week is over, the system goes to its original strictness level.
After every trip, the cumulative points earned is divided by the distance driven by the driver. This is to ensure that all drivers can be evaluated on a level scale regardless of their daily driving distances. This system adds up all the individual trip scores to a cumulative total for the driver which is visible in the mobile application.
The table below lists all the rules, the sensors necessary and the logic behind recognizing the breaking of a rule and scoring points :-
Dashboard visuals and audio feedback
The dashboard display plays an integral role in the attempting to change driver behaviour by providing real-time assists and corrections. While doing this, it should also be conspicuous when not needed. It is important that these corrections be visible to the drivers to enable him/her to drive better, but it cannot be something that will distract the driver from the proceedings on the road. Subtlety was key for this part of the system, given that it was present in the line of sight of the driver.
One of the use cases for the dashboard visuals is when the driver concerned is Tailgating, ie, closely following another car without leaving an appropriate amount of gap between the two vehicles. When the sensors on the car detect Tailgating, the dashboard visual changes to
The severity is denoted by change in colour from yellow to an intense red. The visual of the two cars colliding also provides a cue to the driver as to what the evasive action should be.
These visuals are dynamic and change according to the severity of the “mistake”. They are also accompanied by an auditory cue, beeps. The pace and volume will be used to denote the severity, which will go hand in hand with the visuals.
Likewise, these visuals accompanied by audio cues shall be used to warn the drivers about other mistakes such as cutting and weaving, driving with high beam, driving in between lanes, speeding, etc ( g3).
The App: The Monitoring System
Our mobile phone application gives the user access to all the logged data of each and every trip that he/she has taken. It allows the user to hook his social media profile so that he can see his friends’ progress, view their pro les and learn and get motivated to drive better. The application also allows the user to look into various GR2DC events, badges and achievements and view the car profile as well. The working and the features of the application are explained in the following section.
Journey wise log of points
The journey wise log of points will not only give the user a ledger with all his journeys logged but will also help monitor what mistakes were made and can help recognise patterns on which are recurring problem areas.
Events
Every few days a social event is introduced, the purpose of the event is to encourage the user to consciously follow a certain good driving habit for the specified period of time. Everyone is free to join the event; participation in events will not affect the scoring of drivers, but would give an event badge for attending and the event and following the purpose of the event. A few examples of events include No Honk Sunday, Lane Discipline Wednesday, Low beam Monday etc.
Achievements and Badges
Every driver has a level. As he keeps on gaining point, his level increases. Other than levels he also unlocks achievements as he keeps on driving and improving, for example if the drive does not jump 100 consecutive traffic signals, he unlocks an achievement. The driver also receives a badge for attending events, staying on top of the leaderboard, etc. Badges are majorly for social events but the achievements that a person has are common to every user.
Information Architecture
Future of the Project
In the future, the rst thing to do would be to redefine our center point user and create a new persona, as we realized from our user testing. We realized that this system might work much better for a person to monitor someone else rather than use it to correct themselves. This system might come in handy for a person owning multiple cars that are used as taxis or driven by drivers. Even parents can use this app to monitor whether their child is driving properly or not. We have to redesign the app to suit the needs of the new center point user and persona. We have to make sure that the system works perfectly for them, close any loopholes and iron out any defects present in it. If the system works out and becomes popular, the drivers of personal vehicles will start using this system, ultimately spreading it to 100% of the drivers.
Another change to be made will be to give a url link to a web page corresponding to each driver’s data instead of a mobile app as one may not need to access this information on a daily basis but instead a weekly or monthly basis and in that case the app will end up not being used.
With the advancement in technology in the future, this system will become more accurate and ef cient, using minimum processing and with zero margin of error. Car manufacturers in the future might have all the required sensors and screen already present in the car, thus making the installation of this system as easy as adding a plug-in to the car’s central com- mand center. Car manufacturers can also make this system a standard in their cars to ensure better safety and making its use mandatory for availing warranty on their cars. This system can then be implemented in other vehicles on the road such as Auto-rickshaws, Buses, Trucks and even Two wheelers that will result in 100% of the drivers on the road being disciplined. This will allow for a more efficient and continuous flow of traffic, eliminating congestion totally from the roads.
This project was done as a part of Design for Interactive Media course by Prof. Sudhir Bhatia at IDC IIT Bombay. System developed by Virat Sharma, Raaghav Verma and Shubhankar Biswas.
