BikeWatch: Protecting Pedestrian & Cyclist With Power of AI

Using Machine Learning and high powered Deep-learning Processing Unit (DPU) to detect pedestrians and alert them that a bicycle is approaching, reducing the chances of bike-pedestrian accident.

Biking is a common recreational and commuting activity for many including me. In 2020 alone, there were 52.73 million bikes on the road. With all of these bikes and pedestrians on the road, the probability of accidents has grown in the past year, with 207 pedestrian fatalities in NJ alone in 2020. I wanted to figure out a way in which I could reduce the chances of a bike accident happening on the road.

The solution was BikeWatch. The basic cause of most of these accidents is that pedestrians don’t realize that a bike is coming through until it is too late for both of them to avoid a crash.

Thus, I used state-of-the-art Machine Learning Object Detection Model to detect pedestrians on the road and used a speaker connected to the main hardware to alert the biker and pedestrian to the possible danger of an accident.

The accident can then be averted by the cyclist steering away and pedestrian moving out of the path of the cyclist.

BikeWatch is able to successfully eliminate the danger of accidents when riding a bike.

My solution, BikeWatch, is a device that can be mounted on a bike. The device clicks a photo with a camera and then uses a object detection model to detect pedestrians. It then alerts the pedestrians of the incoming bike. 

Through alerting the pedestrian autonomously of the incoming bike, BikeWatch gives time for the pedestrian to move aside for the oncoming bike and reduces the chance of a bike-pedestrian accident occurring.

The target population is safety-conscious bikers in the United States, as wherever there are bikes there are pedestrians and the chance of an accident. There is no competing product available on the market, and thus my product, which can be mounted on any kind of bike, and is available to any age, represents a blue ocean opportunity for my target population whose needs (advanced bike safety) are unfulfilled.

As both a pedestrian when walking to school and a biker for recreation, I built BikeWatch with an understanding of the needs of both parties. My previous work in making electronics and computer science projects as a maker helped me gain the technical skills to build this awesome product.

As a user of this product myself,when I was designing I made a criteria list as to what I would like in the product:

* Detect pedestrians with a high accuracy

* Alert pedestrians of incoming bike before the bike passes them

* Reduce the chances of bike-pedestrian accidents

BikeWatch is a blue ocean, it fills the needs of an unfulfilled consumer market and has no competitors. It is a game-changing product that can revolutionize bike safety all across the world as a universal device that can be mounted on any bike to reduce the chances of a bike-pedestrian accident and make biking a pleasurable and safe experience.

My goal is to reach out to more and more platforms such as mIT Solve where I can display my invention and innovation with like-minded people and garner support for my products.

The key technology is Artifical Intelligence Object Detection Machine Learning model which is used to detect pedestrians in photo clicked by camera. It is a RefineDet model which is deployed on a Deep Learning Proccesing Unit made to run complex models like RefineDet in real-time. C++ code is used to run the inference and python code is used to take the output from the inference and play the pedestrian alert from the speaker. My product combines many technologies to build a wholesome solution for bikers and pedestrians.

Current number of people you're serving: 1

Number you're serve in next year: As my product is a blue ocean, it could be sold to potential market masses in the next year.

Financial support is needed to expand this product on an industrial scale. Legal barrier is that as a 9th grader in high school it will be a totally new and challenging adventure to start my own company.

None

The BikeWatch device can mounted on any bike and allows safety conscious customers to enjoy biking without the fear of pedestrian-biking accidents. I provide my product at an affordable rate for bikers. As a blue ocean territory, I will not have competition and be able to reap sufficient profit from potential customers due to advantages of biking with BikeWatch.

By selling my product at price of 300$, which will be at 50% margin and can sustain and fund my further R&D efforts in this subject.