IIT Madras students won ISRO’s lunar habitat challenge with an AI that reads moods

SpAr Conclave 2026, a national lunar habitation hackathon organised by ISRO, ANRF, IIT Roorkee and TIDES, asked teams to design habitats for future Moon dwellers.

A habitat that treats mood as life support: an edge AI digital twin reads biometrics and adjusts the environment to astronauts’ cognitive and emotional state.

The four IIT Madras BS students fused AI, neuroscience, architecture and systems engineering into SYNAPSE-1, then defended it before ISRO scientists and IIT Roorkee faculty on structural mechanics, acoustic physics and power budgets.

First place in the national challenge; the win took the team to ISRO’s National Meet on Scientific Exploration of the Moon — three days with Chandrayaan-4 and LUPEX researchers.

For the Space Architecture (SpAr) Conclave 2026 — a national lunar habitation hackathon organised by ISRO, ANRF, IIT Roorkee and TIDES — a team of IIT Madras BS students led by Aparna Jha, with Abhishek Chopra, Pratham Amritkar and Neetika Jain, proposed SYNAPSE-1: a closed-loop cognitive cyber-physical habitat powered by an edge-computing AI digital twin.

The concept starts from a question beyond life support: astronauts on long-duration missions face isolation, confinement and extreme environments, so the habitat itself interprets biometric feedback and adjusts the surroundings to their cognitive and emotional state. ‘We wanted to design a habitat that responds to human emotions and cognitive well-being, not just physical survival,’ Jha said.

Winning meant defending the idea in technical depth — structural mechanics, acoustic physics and power-budget considerations — in front of ISRO scientists and IIT Roorkee faculty. The team took first place, and the win brought three days at ISRO’s National Meet on Scientific Exploration of the Moon, discussing Chandrayaan-4’s sample-return mission, LUPEX (Chandrayaan-5) payloads and human spaceflight under the Vision 2040 roadmap.

Long missions make mental state a hard constraint, so the design treats cognition and emotion as measurable habitat inputs.

An edge-computing digital twin keeps the sensing loop local, without a dependency on Earth round-trip latency.

The team deliberately fused disciplines rarely combined: AI, data science, neuroscience, architecture and systems engineering.

The concept was stress-tested against ISRO evaluation criteria — structural mechanics, acoustics, power budgets — before it won.

Design for the human state, not just survival: the winning lunar habitat treats cognitive well-being as a system requirement, equal to power and structure.

Jha attended ISRO’s National Meet on Scientific Exploration of the Moon and wrote on LinkedIn that she left with ‘a notebook full of ideas’, plus a new interest in low-power machine learning, edge computing and robust data pipelines for space, from robotic payloads to future human habitats. SYNAPSE-1 remains a student concept, an early example of the problems its builders want to work on at the intersection of AI and space exploration.