LightSafe

A common cause of accidents on road during nights, is high-beam being turned on by drivers. And when a driver from opposite direction has brief exposure to this high intensity light, it blinds their vision due to an oversaturation of their retina for short duration, leading to accidents.

A normal driver may not be aware of the state of their headlamp's beam, so it was imperative that we automate this simple task, without hindering their road-vision at any point of time. We provide a simple solution that detects the presence of a car from the opposite direction using a smartphone-camera, to send control-command to an OBD-device that toggles headlamp-beam accordingly for them.

Scaling of a solution becomes simpler when made affordable. We make use of devices that user already has, minimizing their expenses. Being a plug-and-play solution, that's retrofittable into any car (post-1995), makes it easy for adoption too.

Night-driving is becoming risky for most people, rising their levels of stress on the road today. Major factors responsible for this is the inappropriate usage of high-beam on vehicle’s headlamp. The sudden intensity of light blinds the driver for a few moments, leading to most of these accidents. This leads to more than 5000 accidents each month in most cities of India. And this trend is similar in other major cities worldwide.

Different scenarios ...

Worst state:

With both Cars 1 & 2 on high-beam - it blinds both the Drivers 1 & 2

Current state:

Car-1 on high-beam & Car-2 on low-beam - It is safe for Driver-1, but blinds Driver-2 itself

Post-LightSafe:

With both Cars 1 & 2 on low-beam - it is safe both Drivers 1 & 2

This is a problem that effects people worldwide. Most cases of this problem occur due to human lack-of-awareness.

This is a general solution for everyone driving the night. We found that the biggest challenge is making it affordable for all, and retrofittable into all of the cars, can one offer a true solution for this purpose.

Most people weren't willing to spend a lot of money for the safety of others, nor were they ready for major modifications for this purpose, though it meant a lot to them. It is only when everyone uses an automated low-beam solution, can we all benefit from it. 

Having a solution for this on a select few cars too, as it currently is (expensive ones) also wasn't benefiting these drivers, as it was only the drivers opposite to these cars on the road who were getting benefited from it.

To make it easy for adoption, one needs an affordable solution which is easy to install on any car, irrespective of its manufacturer or its year-of-manufacture. Thus we developed LightSafe, which makes use of a smartphone (which most drivers already have) to detect and analyze the presence of an oncoming-vehicle and control the car’s light-beam accordingly. In order to make sure it works consistently, we tested it on different roads, having varied conditions.

Computer Vision Algorithm

For processing the video-stream, we used OpenCV, and programmed using Python. We started by setting up an IP-camera, and it gave access to the captured stream. This output was fed into processing unit where my algorithm runs to detect contours around beaming headlights, which appeared as white blotches.

Contours built on video-stream in the bottom-right-quadrant

Contours built on the cropped image

Camera capture

We used Raspberry-Pi for processing initially, and built complete vision algorithm using that. And once sure, ported it to Android Smartphone for processing with minor modifications - as most users already possessed this device.

During testing, found a few false-positives where system would wrongly interpret presence of a car:

• Tail-light of a vehicle driving in the same direction
• Light from buildings on the road
• Street-lamp on the road

In order to fix them, we detected light coming from the bottom-right-quadrant (for right-hand-drive cars in India), which considerably improved accuracy, to approximately 91% under city-traffic conditions.

Electronic Control

When we realized cars are electronically-controlled, we started exploring possibilities of taking this approach - as it would make controlling car-lights seamless with no moving parts around the driver. All it required was wireless-OBD (On-Board Diagnostics) controller, which could be plugged into a car’s OBD-interface. Smartphone after processing, sent a specific control-code securely over Bluetooth, to switch light-beam accordingly. These control-codes are highly-restricted, so could test this within auto-manufacturer’s office-premises.

Our solution makes use of a couple of core technologies, all integrated together to provide an innovative solution, which is

• Affordable for mass adoption
• Retrofittable into existing cars 
• A simple plug-n-play solution, requiring no major modification of the car
• Real-time optimization using least amount of device resources - this helps to run on varied smartphone device configurations 

All of these factors help it easy for adoption, in order to increase the safety on the road for everyone.

The various phases of the product include:

1. Capture an object/attributes effectively - we do it with couple of vision algorithms to perform a bare minimal processing required.
2. Process the image captured optimally to identify the presence of a car with its light turned on, and not other objects on the road. These are simple fast image processing algorithms for contour detection.
3. Control is the action required, to toggle car's light-beam depending on the conditions in front of the driver at that given moment. This is done by sending out a signal from the smartphone depending on the road status, to the OBD device which actually interprets the code that performs low/high beam on the car.

On the road today we see some accidents happen for no major reason, but rather a simple fault of human negligence. Most people we understood, drive their cars on the road without the knowledge of light beam conditions at all. Some wouldn't even know that they had 2 modes of operating their headlamp. And a few others feel that whatever is good for them, is just good enough to have it. These perspectives of the drivers lead to the fact, that a little bit of care could save so many of these accidents as we have limitations on our own human eye, when a high intensity light falls on our retina.

With this solution, we'd like to address them all seamlessly for them. And without intruding their own preferences. To make our roads a lot safer.

We are conducting our trials. Shall soon be talking to manufacturers and officials from the government, especially the department of transportation and road safety.

Next year it is going to be about creating awareness on road safety, and how this solution can help them in that process. While we conduct trials, we'd like to demonstrate it to various people, and conduct road trials on live vehicles. Using more feedback, we'd like to incorporate changes if necessary. 

During the second-half of the year, we assume there shall be some adoption of this in a couple of major cities of India. And by end-of-the-year we shall be creating this as a defacto feature that most people use when on the road during night times.

India is a huge country, that has this serious problem. Hence to cover the country in full may take an year or two. Following which we shall consider how to spread this across to other parts of the world.

Adoption of this solution by various car manufacturers is the biggest hurdle of all. This solution is very dependent on the engine control codes, which aren't open to the public for reasons of security. Hence, their cooperation for this is vital.

Another aspect is about creating the awareness of a solution for road safety in a country where most people still do not understand much about it.

Urban cities we presume could be easy to convince, but still needs an awareness campaign driven by the government for this purpose. And this could take a while, as that process needs a lots of patience at each stage.

We are actively talking to Original Equipment Manufacturers of the car who provide the required auto electronics for the car manufacturers. This could help us in the first stage before we start any discussion with the actual manufacturer itself for an official understanding with them.

Rishank Kanaparti

Primarily it is our software skills, to help build an app on a device. And the domain knowledge in auto electronics.

None, officially. 

Currently, we only interact with friends from the car OEM industry. And a few others who run NGOs related to road safety.

The solution is primarily focused on providing road safety for everyone, especially those driving during the nights. This being a huge problem, it caters to a mass market. For this purpose, we shall be working towards building a resilient solution that can securely be deployed by various car manufacturers, as a certified accessory device on their vehicles. Also in order to spread its public awareness, we would need cooperation from transportation department of the government.

We’d like to drive this product as a contribution to the development of a nation. As a technology provider, we would like to keep it that way. With minimal funds required to sustain its development. The royalty fee charged shall be in proportion to the total cost of the product when it is released into the market.

Solve can help provide us the necessary awareness to this huge problem, not only to the various stakeholders who are involved to help solve this - such as government, automotive manufacturers, and of course the people on the road who are the users finally. With a small team having limited resources and reach, which takes a lot of effort. Efforts could rather be diverted towards the development of the solution. Also appreciate the generous funding one can receive from this platform, which can help better execute this solution, in a more effective way.

Original Equipment Manufacturers for the automotive companies, such as Bosch, Continental, Delphi, and many more.

Automobile manufacturers such as GM, Suzuki, Ford, Honda, Volkswagen, and many more.

Transportation department of the government of India. Eventually similar departments in other countries.

Our solution shall eventually incorporate a lot AI into its systems, as the devices too will have the necessary intelligence to support it. AI basically in order to learn and take better actions on the vehicle, depending on the driving habits of its user, road conditions in which they drive in, etc.

GM Prize shall greatly help us develop this product very much. Being a solution for the automobiles, this prize would be quite apt too, validating the support from the leaders in this industry.

Morgridge Family Foundation will be a great boost, to bring about a complete shift in the way we think about safety on the road with a solution such as LightSafe.