DEV Community

Cover image for Smart India Hackathon 2023 winners
Dinesh A
Dinesh A

Posted on

Smart India Hackathon 2023 winners

Silence enveloped the room; all eyes focused on the podium, and every ear eagerly awaited the results. The announcement finally came: the winner of SIH1370 was Team DORA the Explorer. The moment those words echoed, joy and excitement filled the air. It was a remarkable feeling to become a two-time SIH winner.

Hello, everyone. I am Dinesh, and my team emerged victorious in the Smart India Hackathon 2023 software edition. We chose to tackle the problem statement "Real-Time Monitoring of Infrastructure" provided by the government of Jharkhand.

Why did we participate in SIH?

We participated in the Smart India Hackathon to learn and validate our entrepreneurial skills. The experience of forming a team, deciding on product features, and making it commercially feasible with a better user experience is invaluable. We wanted to build something that would create impact and solve real world problems. Participation in initiatives like SIH lays the foundation for later entrepreneurial ventures.

How did I form a team?

To build a well-rounded team, I sought diversity in both technical and non-technical skills. Eventually, I formed a team comprising Ishika Jain, Sidhanti Patil, Namrata Hakari, Gagan S, and Prince Thakkar. Our team covered designers, business developers, ML engineers, full-stack developers, and presenters. We named our team as Dora the explorer
Dora = Dynamic Observation Real Time Analysis.

How did we choose a Problem Statement?

We selected a problem statement aligned with our personal goals and within our skill set.
We had to choose one problem statement out of 234 problem statements. Criteria for choosing included: no dependencies, standalone solutions, fewer competitors, government not being the sole customer, monetization potential, large Total Addressable Market (TAM), and no hardware equipment's. Problem statement SIH1370, Real-Time Monitoring of Infrastructure, met these criteria, and we chose it to work on.

How do we approach a problem?

Our unique approach was that we didn't immediately jump into solutions or coding. Instead, we followed a design thinking strategy, spending 60% of our time understanding the problem, defining user needs, ideating solutions, prototyping, and testing.

Our Solution :

Logo of Pratyaksh

Our solution was built on an AI and GIS foundation, providing real-time monitoring capabilities to construction sites. Key features of our solution included:

  • Progress Tracking:

    • Utilizing AI algorithms, we implemented progress tracking by comparing current images of construction sites with the previous day’s images captured through CCTV cameras.
    • Material consumption at the site was also calculated through CCTV cameras to provide an additional metric for progress.
  • Project Management System: We developed an end-to-end project management system allowing contractors and project managers to create milestones, tasks, and assign them to site engineers with due dates and priority levels.

  • Mobile Application for Site Engineers: To facilitate daily progress updates, we created a mobile application for site engineers. Through this app, engineers could upload daily progress reports and mark the photos of completed work on the site plan.

  • GIS Mapping: All projects were mapped on GIS, enabling anyone to view the current status of any infrastructure project and analyze various patterns.

User Flow

How did we validate our idea and solution?

We pitched our idea to various stakeholders, gathering feedback from friends, faculty, contractors, civil engineers, and ML specialists. This feedback helped us identify and address potential issues and refine our solution before the grand finale.

How did we choose a tech stack?

When selecting our tech stack, we prioritised a large community for support and ensured at least 50-60% familiarity within the team. Our tech stack included ReactJS, NodeJS, Express, MongoDB, React Native, Leaflet, Flask, and various ML algorithms.

What happened in the Finals?

Our picture at Grand Finale
Our finals were on December 19th, and our nodal centre was in Jaipur, Rajasthan. Travelling with my teammates was a fun experience filled with singing, playing, joking, and final preparations.

The finals comprised three judging rounds and two mentoring sessions.

The first judging round focused on presentation, uniqueness of the idea, and feasibility, accounting for 20% of the total weightage.
Judges emphasised inclusivity, urging us to ensure that no user could find fault with the product. They also set challenges for us to include air quality monitoring and make the product adaptable by integrating with other services. This round we did pretty well.

The second judging round focused on the tech and execution , accounting for 30% of total weightage.This round didn't go well as we misunderstood the term integrating but I guess we scored well as the round focused on execution. Judges challenged us to integrate various components, including ML models with the application, connect frontend with backend, and dynamically map projects on GIS for the final round.

The third round was forced on complete end to end product, commercial viability and end user perspective, accounting for 50% of the total weightage.
Guess what the application broke just one hour before the last round while integrating, we got multiple merge conflicts, finally we integrated by reverting code and implementing the latest changes by storing the code on WhatsApp messages.
We panicked while pitching at finals, and got stuck here and there, but everyone explained their parts which gave unexpected brownie points as all the team members were involved in the pitch. We put our end to end implementation plan on BMC, which also earned brownie points

And the rest is history.

Read about my SIH'22 journey here.

Top comments (0)