Gokul NA on training robots to learn like infants: CynLr’s Object Intelligence Stack

In today’s episode, we dive into the future of robotics with Gokul NA, founder of CynLr, or Cybernetics Laboratory, perhaps one of India’s most advanced companies in this field.

CynLr is headquartered in Bengaluru, with an advanced R&D Lab in Switzerland and a growing customer-facing operation in the US, where Gokul’s fellow founder Nikhil Ramaswamy is based.

The entrepreneur duo and its 85-member team are tackling a challenge that scientists and engineers have been working on for several decades: the ability for robots to intuitively handle unfamiliar objects without custom programming or prior training.

In this conversation, Gokul explains their new “Object Intelligence Stack,” a system designed to imitate some of the functions related how a human brain might learn — much like a baby impulsively grasping a new toy without knowing its name or purpose.

By collaborating with the Centre for Neuroscience at the Indian Institute of Science, CynLr is translating brain-function research into a sophisticated software and sensor framework. Shifting the focus from algorithms that work on data to the “physics of objects,” CynLr’s intelligence stack is what Gokul describes as the precursor to a “manipulation OS.”

Drawing parallels to Apple in the 1980s, Gokul shares his vision for a future of “object computers” and micro-factories – important components of CynLr’s vision for global manufacturing, where instead of Giga-factories, we might have fabrication facilities as small as a car dealership or even a garage.


Chapters

(00:00) Challenges of building a deep tech organization in an absent industry

(05:58) Imitating the human brain’s ability to handle unfamiliar objects

(09:18) Partnering with neuroscience researchers to replicate human intuition

(11:32) Developing a manipulation operating system and the future object store

(19:47) Automating assembly for automotive and semiconductor manufacturing

(25:17) Transitioning from rigid gigafactories to software-defined micro-factories

(35:49) Fostering a deep tech ecosystem to address the brain drain

(40:18) Strategic funding goals and the technical roadmap for scaling


Gokul and Nikhil are backed by investors including Speciale Invest, growX ventures, Pavestone VC, Athera Venture Partners (formerly Inventus India), Anicut Capital, Arali Ventures, Redstart Labs, and several other institutional and angel investors.

CynLr’s long-term vision also involves creating a manipulation operating system and an “object store” and a “task store” to transform flexible manufacturing, just as the App Store transformed the smartphone.

Currently, they are deploying these solutions in the automotive and semiconductor industries to automate some complex manual assembly processes.

In this conversation, Gokul also talks about some of the challenges of building a company like CynLr in India, where many important ingredients are missing, and what he thinks the industry can do to change that.

Listen to the episode

Comments

Leave a Reply