Smart India Hackathon 2020

“Everything around you that you call life was built by people that were no smarter than you. And you can change it, you can influence it… Once you learn that, you’ll never be the same again”

-Steve Jobs 

Smart India Hackathon is an annual nationwide innovation-driven competition for college students and held separately for school students by the Government of India. Numerous ministries, institutions, and PSUs under the GoI, private corporations from India, and various state governments participate in this hackathon to propose problem statements relevant to their institution. This hackathon takes place in 2 separate editions: hardware and software. And for SIH 2020, 145 hardware problems and 344 software problems were proposed by various institutions and companies for the young scientific minds to work out and innovate. The winning team from each problem statement is given prize money of Rs. 1 Lac along with the opportunity and support to propel their innovation into a startup. The winning team, if they choose to proceed with a startup, is incubated by the team YUKTI, which works under the Ministry of Education’s Innovation Cell. This year 2 teams have brought laurels to our institute. Team O.S.C.A.R. won in the problem statement proposed by National Mission for Clean Ganga, building an automated robot to detect garbage and clean the river Ganga, and Team Numero Uno won in the problem statement proposed by the Ministry of Housing and Urban Affairs focusing on software-based disaster management solutions for smart cities.

Team O.S.C.A.R.

  • Abhay Shukla
  • Sridhar Singhal
  • Rutav Shah
  • Nikunj Bagaria
  • Radhika Patwari
  • Raushan Prabhakar

Team Numero Uno

  • Aditya Parashar
  • Krishnam Kapoor
  • Karan Mahajan
  • Amrit Goswami
  • Tiyasha Mitra
  • Hariharan Ravichandran

We talked to both the teams about their journey from ideation to victory and their projects’ future plans. Their stories can potentially spark many KGPians to go forward with their idea, build a team and probably get into business. 

The format of the competition is the following. First, a team presents its idea, design, and/or prototype in the internal hackathon/selections held in college. Each year a prescribed number of teams are selected to represent the college in hardware and software editions. In the next stage, the submitted technical reports from all over the country are screened by the SIH team. Only 5 teams from each problem statement proceed to the finale, where rigorous mentorship and presentations are scheduled. 

The solution that Team O.S.C.A.R came up with was initially built by the boarders of Rajendra Prasad Hall of Residence for the Hardware Modelling event of General Championship Technology 2018-19. For that event, the team had built a working prototype capable of remote-controlled traversal on the water with a mechanism to collect garbage from the surface. After the release of the problem statement from SIH, the team, in view of automating the bot, which was only remote-controlled till now, invited more members having an inclination towards the software end. Since the mechanical construction and the electronics were already nearly finished in 2019, they planned to work on automation during the summer vacation of 2020, but then the pandemic hit. So during the summers, the team did all the static analysis of the mechanical structure and electronics to present the durability and reliability of the bot in front of the jury. Automation stack comprised of many modules sequentially dependent on each other. Firstly, for machine perception, a DL model was used to detect garbage while relative position from the bot was computed in closed form. Once the positions of garbage and obstacles are known, an optimal path is planned, and then the bot has to be precisely controlled to follow the path smoothly. Hence the perception, planning, and controls module was built during the summer and was tested to some extent on simulators. All these results were presented before the jury. The accuracy of all the results, mechanical design and reliability, safety measures of the bot(like waterproofing), and potential efficiency brought the team victory. They plan to integrate the software stack in the bot, make some improvements and test it once they reach the campus.

For the software edition, the problem statement targeted by Team Numero Uno was also targeted by 26 other teams from India. The format of the competition is slightly different from the hardware edition in the way that here the final round is a 3-day hackathon where continuous improvements are made based on the feedback from assigned mentors and presented before the jury periodically. Few members of the team selected this problem statement of ‘disaster management in smart cities’ while looking for one with the scope of applying ML in the solution. More members joined the team to build a full-stack system contributing to the user interface, backend, etc. In their disaster management solution, the system is comprised of 3 modules that function in a sequence. The first is the detection module that uses a huge network of CCTV cameras laid across all the streets in the smart city. A deep learning model was trained to detect and classify disasters like fire outbreaks, riots, flash floods, crowd accumulation during the pandemic, landslides, etc., all based on the video feed from the CCTV cameras. Once the region and class of outbreak are identified, the next module sends alert messages to residents in nearby regions depending on the outbreak using a closed-form model. Then the most crucial step is to optimally and efficiently supply relief material to the affected sites and evacuating people from there to the assigned safe zones. The team used a route optimization algorithm, based on research published by one of the team members, to compute an optimal plan for the relief squad depending on the location of the affected region, relief material storage, and safe zones along with other variables like the capacity of transportation facilities. These three modules enhance the disaster management routine in every way. For the final round of the competition, they demonstrated a prototype with all the aforementioned features for fire outbreaks, flash floods, crowd accumulation, and landslides. The robustness, efficiency, concrete mathematical support, and holistic approach brought them a lot closer to finally winning the competition in their problem statement. 

This goes without saying that the nation expects the students from elite engineering colleges to build robust, economical, and efficient solutions for the problems faced by various demographics, corporates, or institutions. More importantly, students must always find ways to put their learning into practice at the largest scales possible and feasible. But not everyone pursues this direction during their UG years. Smart India Hackathon gives the opportunity to work on both fronts, with full throttle, in a team. Not just it solves the problems proposed by various institutions, potentially to be deployed by themselves it also decorates the team with recognition and huge prize money. Also, it gives them the opportunity and support to get incubated into a startup. KGPians performing exceptionally well every year in SIH inspire more undergrads to pursue this direction of learning via innovative projects, teamwork, and solving large-scale problems. Awaaz, IIT Kharagpur congratulates all the teams that participated in SIH 2020 in various problem statements, giving their best to solve them despite the pandemic, irrespective of their final standing in the competition.

Leave a Comment

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