Pre-Final undergrad at IIT Jodhpur, studying Mechanical Engineering with a specialization in Robotics & Mobility Systems.
I spend most of my time trying to get robots to understand the world, teaching them to see where they are, feel what they're touching, and move like they mean it. Somewhere at the intersection of perception, manipulation, and reinforcement learning is where I live.
Currently obsessed with making robots less clueless in the real world. Point clouds, force sensors, sim-to-real gaps : the whole circus.
When I'm not debugging why my robot thinks it's somewhere it's not, I'm probably reading a paper that'll make me rewrite everything I just wrote.
I care about:
- robots that can localize themselves without losing the plot
- manipulation that doesn't fall apart the moment geometry changes
- models that transfer from sim to reality without crying