Hackathon
Bishrant Ghimire

Tombs, a Blown Tyre, and a Best Emerging Team Trophy

My first hackathon experience at DELTA 5.0 in Dharan, building an L2G fault detection system with relay modules, nearly getting hit by a truck, and winning Best Emerging Team.

HackathonHardwareTeam WorkCompetition

This is the story of my first hackathon. It involved a relay module, a scooter that nearly got us killed, and a guest house completely surrounded by graves. I am not making any of that up.


Dharan

DELTA 5.0 was held at IOE Eastern Region Campus in Dharan. I'd never been. The city sits at the base of the Siwalik hills in eastern Nepal and there's something about it that feels genuinely different from Kathmandu. Quieter in a specific way. The kind of place where you notice the sky.

We arrived to clouds. Not dramatic storm clouds, just the heavy, low, grey kind that sit over the hills and make everything feel slightly cinematic without actually being cinematic. It rained on and off. The campus grounds were damp. The whole experience had this texture to it that I keep associating with the word "first" because it was.

First hackathon. First time competing in something like this for real, with a team, with judges, with something to actually lose.


Hariyaali

We needed somewhere to sleep. The place we found was called Hariyaali.

Hariyaali was fine, genuinely fine, except for one detail that nobody mentioned when we booked it: it was entirely surrounded by tombs. Not like, there was a cemetery nearby. I mean we were in it. The guest house sat inside what was essentially a burial ground, graves on every side, some of them right up against the building.

We discovered this at night obviously.

I don't know who decided this was a good location for a hospitality business but I respect the confidence. The rooms were decent. We slept fine. But walking out at 2am to get some air and realizing you're standing in a graveyard is an experience I will carry with me.

Adhish thought it was funny. I was mostly trying to decide how I felt about it.


The Scooter

This is the part of the story that, in retrospect, I cannot believe actually happened.

We were getting around Dharan on a scooter. Four of us, logistics unclear, the kind of thing that makes sense in the moment and sounds insane when you describe it later. At some point the tyre exploded. Not a slow leak. The tyre just went, sudden and loud, while we were moving.

What happened next I can only describe as Adhish being very good at controlling a scooter. The vehicle started going sideways toward oncoming traffic, specifically toward a truck, and then it didn't because Adhish corrected it somehow. It happened fast enough that I didn't fully process it until we had stopped and were standing on the side of the road.

There was a truck. We were fine. Those two facts coexist in my memory and I'm still slightly amazed by it.

We stood there for a moment. Then we sorted out the tyre situation and kept going. That's how it goes with first hackathons apparently. You nearly get hit by a truck and then you go back to thinking about relay logic.


The Project

The theme was Technology in Agriculture and Innovation. We had 12 hours. Our team was me, Shishir Poudel, Adhish Paudel, and Mausham Sigdel.

We built a single-phase line-to-ground fault detection system using a relay module as the main component.

The idea was grounded in something real. In agricultural contexts, especially in places like rural Nepal, electrical faults in single-phase distribution lines are common and often go undetected until equipment is damaged or someone gets hurt. A line-to-ground fault happens when a live conductor makes contact with the ground, either through insulation failure, physical damage, or environmental factors. Detecting that fault quickly and isolating the circuit matters.

Our system worked by monitoring the current waveform in the distribution line. Under normal conditions, the current profile is predictable. A line-to-ground fault creates a characteristic disturbance, a sudden surge and asymmetry, that the relay module can be configured to detect. When the fault signature appears, the relay triggers and isolates the faulty section.

The relay module was the heart of it. We configured the relay logic to respond to the threshold conditions that indicate a fault while ignoring normal load fluctuations. Getting this threshold right was the interesting part, too sensitive and you get false trips, not sensitive enough and you miss actual faults.

12 hours is not a lot of time to build, test, and present a hardware system. We spent most of it at the bench, wiring, testing, adjusting. The last stretch before the presentation was the usual kind of controlled panic where everything is mostly working and you're trying to make mostly into completely.


The Part We Didn't Expect

We won Best Emerging Team.

I want to be honest about this: I didn't expect that. We were first-timers. The project was relatively simple compared to some of what else was being presented. We weren't doing anything with machine learning or advanced sensing or complicated software.

We were doing something useful. I think that's what the recognition was for. A practical solution to a real problem, built from scratch in 12 hours, by people who hadn't done this before.

The feeling of winning something you didn't expect to win is strange. Not bad strange. Just the kind of thing that recalibrates what you think you're capable of. We showed up not knowing how this worked, nearly didn't make it in one piece thanks to a scooter tyre, slept next to graves, and came home with a trophy.

That's a pretty good first hackathon.


What I Took Away

Hardware under deadline is genuinely its own discipline. You can't just iterate as fast as you can with software. When a physical system doesn't work, the failure modes are different and the debugging is slower. You learn to be more careful upfront because fixing things later is more expensive.

I also learned something about what competition judges actually value, which is not always what you assume going in. A polished, well-explained solution to a real problem can hold its own against more technically complex entries.

And I learned that Adhish is a very good scooter driver. That one came up fast and unexpectedly but I'm glad I know it.

Dharan, cloudy and damp and beautiful in its own way, will always be the place where this started for me. I'm going back.