Building Robots with Perception and Purpose -Raviteja Chivukula - Perceptyne |Episode 4|BotCast|AIRA
Raviteja Chivukula, co-founder of Perceptyne, discusses industrial robotics, the company’s dexterous robot platform, and its plans for AI-driven factory automation with the Bot Cast - By AIRA host.
In this talk
- 1:22
Why Perceptyne focuses on hands and perception
Chivukula describes Perceptyne’s industrial humanoid approach: prioritize what robots do with their hands and eyes over walking, since factory floors are designed to be flat. The team is building a dual-arm system for assembly and object handling, where visual and tactile feedback support dexterous work.
- 6:35
A reusable platform for automation
Chivukula says Perceptyne aims to bring the scale and adaptability associated with SaaS to robotics. The founders, who had more than a decade of hardware experience, evaluated customer demand before committing. Their goal is a broadly useful platform, rather than a robot designed for only one narrow task.
- 32:19
How PR-34D learns assembly tasks
Perceptyne’s PR-34D has two seven-degree-of-freedom arms and three-fingered hands with tactile sensing. An operator can teleoperate it to demonstrate a task around 50 to 100 times, while the system records vision, force, touch, and joint data. The company combines imitation learning and computer vision models with its own task-planning work.
- 27:48
Where dexterous robots could add value
Chivukula argues that conventional automation works well for high-volume, low-mix production, but can be hard to justify for variable assembly work such as installing phone, car, or electronics components. Perceptyne is pursuing electronics and automotive manufacturing, with early proof-of-concept work underway in India and abroad. The company has a goal of deploying 1,000 robots and expects broader applications to follow as the technology matures.
- 1:07:26
Building the company and its talent base
Perceptyne has raised pre-seed funding from angels and venture investors, including Bhavish Aggarwal, Tarun Mehta, and Swapnil Jain, alongside government support and incubation at IIT Madras. Its roughly 20-person team spans mechatronics, AI, and applications. Chivukula says robotics talent is harder to find than general software talent, and cites a fresher salary around ₹9 lakh.
Briefs summarise talks and interviews; timestamps open the video at that moment. Claims in a talk are the speaker's own.