Featured project

NexBot — AI Humanoid Robot

GENBOT's flagship humanoid: adaptive learning, modular architecture and real-time environment response.

NexBot is the flagship project of the GENBOT Robotics Club at IARE, Hyderabad. It is an AI-driven humanoid designed to learn from its environment and respond to people in real time. NexBot also lives on this website as the 3D concept animated in the page backgrounds.

Objectives

  • Build a humanoid robot with adaptive learning — behaviour that improves from real-time environment data.
  • Keep the platform modular so mechanical, electrical and AI teams can work on NexBot independently.
  • Give the robot a vision pipeline for detecting and interacting with humans.

Design overview

The project is split into mechanical, embedded and AI workstreams. The architecture is deliberately modular: the locomotion, perception and learning submodules each expose clean interfaces, so improvements in the model or the hardware do not block each other.

How the AI works

NexBot combines on-device inference with adaptive learning. Sensor and camera streams feed a perception layer that builds a real-time model of the environment; control decisions are computed on the robot. Because the system runs locally, response latency stays low enough for natural human interaction.

Project status

NexBot is GENBOT's featured build of the 2025–26 season. The team is currently iterating on the vision pipeline for human interaction and the modular actuator architecture.

Related work

  • DroneX — shares the computer-vision and edge-processing stack.
  • Gesture Desktop — hand tracking research that feeds into human interaction.

Core technologies

Adaptive learning · on-device inference · modular electronics · vision-based interaction

Status

Active flagship project — 2025–26 season

Want to build the next NexBot?

AI and robotics project teams at GENBOT are always open to new members.

Join GENBOT