Microsoft Learn AI Skills Challenge Pitch Winner: Smart Traffic Light System

Posted by

This post has been republished via RSS; it originally appeared at: Microsoft Tech Community - Latest Blogs - .

The Microsoft Learn AI Cloud Skills Challenge wrapped up an incredible learning journey with the AI pitch Challenge; a showcase of innovation where passionate learners brought their visions to life through the power of AI. These creators shared how they would harness Microsoft's AI technology to craft solutions for the future in a 3-minute video pitch. Out of many, five outstanding winners emerged, each with a unique and compelling vision.


This series of blog posts spotlights each creator sharing the transformative potential of their ideas. 


I’m Nischal Sharma, a computer science engineering student from India. During my summer break I joined the Microsoft AI Skills Challenge 2023, The journey helped me learn many in hand AI builder, one of it was Azure Computer Vision.  Which instantly struck my mind to solve a problem I experienced in my daily life, long waiting queues at traffic light and no passage being given to the emergency vehicles at the traffic intersections. Using the insights of computer vision I learned during the challenge, I pitched the idea of smart traffic light management system. Which, unlike a normal traffic signal would analyze the amount of traffic incoming from each direction and give the appropriate amount of green time to the busy lanes or the lanes with emergency vehicles. 



Uncovering the inspiration behind the innovation 

(source: The Times Of India article)(source: The Times Of India article)


As a daily commuter traveling from home to university, I've often found myself frustrated by the long queues at traffic lights, especially at intersections leading to hospitals. Ambulances, with lives in the balance, are forced to wait as impatient drivers block their passage. It's during these early morning hours, when you expect to reach your destination quickly, that traffic jams and imbalanced traffic patterns are most infuriating. My first encounter with imbalanced traffic was about five years ago when my family was running late for a gathering. While three lanes saw the traffic cleared within half the time, our lane waited for the light to turn green for the fourth time. This imbalance not only causes frustration but also contributes significantly to pollution levels.  


Studies indicate that pollution levels at traffic intersections are 29% higher than on open roads, with 25% of harmful particle exposure occurring during the mere 2% of the journey spent passing through intersections with traffic lights. 

Every time I witnessed such situations, I wondered if there could be a better way to manage traffic, one that eliminates the need for traffic police to stand under the scorching sun. However, it wasn't until I explored Azure Computer Vision during the AI Builder Challenge 2023 that I realized the solution was already within reach.


Illuminating the 'Smart' in Traffic Lights 

In my project, I used Azure AI Vision, trained on diverse vehicle data, to create a smart traffic light management system. It uses live CCTV footage to track traffic and optimizes signal timing based on specific scenarios. The system uses, Spatial Analysis and Object tracking services for its operations.





Skills Challenge- a learning milestone 

"AI is the future," 
"AI will change the world," 
"AI will bring revolution"  
These statements have echoed for years. Yet, for non-technical individuals, they remain mere abstract phrases, akin to watching a magician pull a rabbit from a hat. Microsoft Cognitive Services, however, bridges this gap, allowing individuals from any background to not only experience the true essence of AI but also deploy their AI solutions with unparalleled ease, enhancing their productivity. 
For me, the Microsoft Learn AI Skills Challenge was the catalyst that transformed my problem statement into a tangible solution. Initially, all I had was a challenge statement – like a painter wandering with a blank canvas. However, the Challenge provided me with the confidence that the change I envisioned for the conventional traffic management system was achievable and wouldn’t be difficult with the support of computer vision. 


Guided by Microsoft: A Transformative Interaction 

During this journey, I had the opportunity to engage with Microsoft Cloud Advocates. Their invaluable insights and constructive feedback significantly enriched the Smart Traffic Light System project. A notable meeting with Pablo Lopes and Jasmine brought forth the importance of customizing the solution to specific regions by understanding the unique ‘why’ behind the problem. This meeting marked a significant milestone in our project, and their insightful suggestions have motivated us to refine our approach further. 

Jasmine’s contribution was particularly noteworthy, providing crucial guidance on the future trajectory of this project. Her insights are set to influence our ongoing development efforts. The interaction with Microsoft’s product managers not only fortified our project but also underscored the collaborative spirit of our AI journey. 


Next Steps are that of a Pedestrian 

As we move forward with the development of the Smart Traffic Light System, I'm committed to making it not only efficient for vehicles but also safer and more convenient for pedestrians. One of the key enhancements on our roadmap is the integration of pedestrian-friendly features. We will leverage the same Azure Computer Vision technology to monitor pedestrian flow at intersections.  

  • Pedestrian Detection: Implement real-time pedestrian detection using computer vision. This will ensure that the system can identify and count the number of pedestrians waiting to cross the road. – 
  • Dynamic Crosswalk Timing: Based on the number of pedestrians waiting, the system will dynamically adjust the signal timing to prioritize safe pedestrian crossings. If a significant number of pedestrians are waiting, the system will extend the green signal to allow them ample time to cross safely.  
  • Prioritized Crosswalk: If a senior citizen is detected then the pedestrian light won’t turn red unless they have crossed the road.  

Pedestrian safety and convenience will be at the forefront of our development efforts. I'm excited about this new phase of the project, as it aligns with our mission to make traffic management not only more efficient but also more human-centric. Together, we can create a smart city environment where everyone can move around safely and efficiently.  


What Was Your Favorite Microsoft AI Technology to Learn About?  

Throughout the Challenge, I had the opportunity to explore several Microsoft AI technologies, but one that truly stood out was Azure Computer Vision. Its capabilities in image recognition and analysis opened up a world of possibilities for me. I found it incredibly versatile and user-friendly, which made it a valuable tool in developing the Smart Traffic Light System.  


What's the Next Thing You Want to Learn About AI? 

AI is a continuously evolving field, and staying updated is crucial. In the near future, I'm eager to delve deeper into reinforcement learning and 3-d model generation form a 2-d image. My desire stems from their immense potential to create smarter, more capable AI systems to tackle complex real-world problems in novel ways. These skills will not only enhance the capabilities of my current project but also open doors to exciting opportunities across the AI landscape. 


In conclusion, the Microsoft AI Skills Challenge 2023 has been a transformative journey. It not only allowed me to apply AI concepts to solve a real-world problem but also gave me the confidence to pursue more ambitious AI projects in the future. I'm grateful for the opportunity and look forward to a future where AI-driven solutions like the Smart Traffic Light System can enhance our daily lives and contribute to a safer, more efficient world. Thank you for joining me on this journey, and I encourage everyone, regardless of their background, to explore the incredible world of AI through Microsoft's AI Builder and Cognitive Services. Together, we can harness the power of AI to shape a better future for all. 


Feeling inspired? The Microsoft Learn AI Skills Challenge may have ended but the learning never stops! Get started with an AI Learning Path and find a new Microsoft Learn Cloud Skills Challenge to join. Transform your innovative ideas into reality with Azure credits through the Founders Hub. And for the students who dream of making an impact, the Imagine Cup is currently underway!

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.