Inter IIT Tech Meet: The Silver Syndicate of “DRDO DGRE’s Vision-Based Obstacle Avoidance Drone”
The Inter IIT Tech Meet is a technological competition held among the 23 IITs. This year, the competition was organized by IIT Guwahati in a completely virtual mode. The theme for this year emphasized building towards a self-reliant, progressive India with the aim of providing solutions to rejuvenate various important economic sectors such as Agriculture, Tourism, Retail, Finance, and Automobile which have suffered tremendously due to the ongoing pandemic.
The various competitions and events were classified broadly into 3 categories- High Prep, Mid Prep, and Low Prep events.
Among the Mid-Prep events was the ‘DRDO DGRE’s Vision-Based Obstacle Avoidance Drone’ event. This event was set on the backdrop of the recent flash floods in Uttarakhand and other such disasters in remote and difficult-to-access areas. The problem statement involved designing a complete autonomous aerial vehicle capable of detecting and navigating through various static and on-field obstacles to reach its required destination. Such drones will prove extremely effective in negotiating tough terrains and environments and will be key in search and rescue operations.
The team from IIT Kharagpur, led by Captain Praneet Jain (a 5th year UG student from the Department of Agriculture and Food Engineering) bagged the silver medal (joint runners-up with IIT Guwahati). On conversing with Jaskaran Singh Sodhi (third-year undergraduate from the Department of Manufacturing Engineering), we discerned the following:
For the selection of the team, a post was shared from Technology Gymkhana’s Facebook page with a form for interested students to fill up and submit along with their resume. After shortlisting resumes, the students were finally selected after a short interview round.
Successful participation in the event required knowledge of topics such as AruCo detection, computer vision, image processing, ROS, etc. Attending the Image Processing Winter Workshop would be really beneficial to get an exposure to the basics, and being a part of any of the research groups (especially AGV) would certainly help to get a grip over the conceptual understanding as well as the practical working. There are many resources available online about path planning algorithms that may be referred to by anyone interested to proceed in this field.
Talking about his experience, Jaskaran commented, “During the initial days, some members were assigned to the landing part while others worked on exploration and planning. These initial days consisted of literature review and exploration before we started with the actual algorithms. After this, as the subtasks were identified, teams were made accordingly and it was ensured there was always a senior available to help us out. The captain made sure to keep track of the progress of every single team member so that everyone had enough work to do and no one was over-burdened so as to not hamper their academics.”
The virtual mode has derailed various activities in the wake of COVID-19, and even the Inter-IIT squad wasn’t spared of its wrath, as corroborated by Jaskaran too, “As a team event, coordination and communication are the most important factors and with the team members at their homes, this was a challenge. However, regular meetings, regular updates being shared on a messenger group, and the tasks being properly coordinated by the Captain ensured the team dealt with this hurdle effectively. The team members were also encouraged to join a Google Meet and work together. This not only helped in developing camaraderie among the members but also made sure everyone knew what work was happening.”
About the journey to clinch the silver medal, Jaskaran mentioned, “The team was posed with a tough scenario when we had to decide between the 2 algorithms- Vision-based Segmentation method and Global Planner method to be used for the drone. The main predicament was that the Segmentation method was extremely fast (at least 3–4 times faster than the Planner method) but had inaccuracies that would cause the drone to fly around in circles at times. To tackle this, we decided to devote a few more days to this process and worked towards testing and optimizing the parameters for both algorithms. The conclusion that came out of this period was that the Segmentation approach, though quicker, was only 90% accurate, and aiming for the medals, we decided to go forward with the Planner method. They say, failures are stepping stones to success. The team collectively worked through all the failed attempts and finally made all the crashes worth the effort. This was my first Inter IIT and this experience was truly enlightening and rewarding and surely the highlight of my second year. A certain upper hand such competitions have is they require you to learn topics that you might not experience in a research project. The high stakes push you to work unimaginable hours and learn things faster than normal.”
For those who strive to emulate this path and dream incessantly about being a part of the Inter-IIT Squad for events related to robotics, Jaskaran advised, “For those just starting out exploring the field of robotics, familiarize yourself with the curriculum of both winter workshops as both are equally important in this field. They serve as a great starting point. After that, getting some practical knowledge through different kinds of projects will help you a long way. To get into the teams, some basic knowledge about the field of whichever event you wish to get into is required. And in terms of the language requirement, we worked both on Python and C++, and in interviews as well, emphasis was laid on concept clarity and experience with some robotic algorithm along with zeal and interests of candidates rather than language-specific knowledge. Having started exploring the fields of robotics and computer vision towards the end of my first year, I didn’t really expect to get into the team, but turns out, being part of the team doesn’t require you to be a big-shot roboticist. It just requires curious and hard-working people. People who made the effort to explore the field and want to apply the skills they possess. And so my advice to juniors is to just explore the field, not for the sake of getting into the team, but to learn more and more. And when the time comes, put forward your application, and let your work speak. Lastly, and perhaps most importantly, it does not matter if you’re in a research group or not. All that matters is your curiosity, and the amount of effort you’re willing to put in. There were numerous people in the contingent who weren’t a part of any research group but did wonders.”
The Scholars’ Avenue congratulates the entire team for their splendid performance and wishes them luck for their undertakings in the future.