Necessity is the Mother of Invention: How We Went from Building a Robot to Building a Supercomputer

Veerbhan Kheterpal
May 14 · 4 min read

Co-Founder and CEO of Quadric.io

Our story was going to have a very different ending. In 2016, we started building an agricultural robot that was going to transform first, vineyard management, and eventually, crop management of any kind — reducing costs, minimizing arduous tasks for humans, and maximizing crop returns. While this might seem like a lofty goal, I was working with two talented partners, and we had just built and shipped a very complicated technical product in the bitcoin computing space. With our combined technical backgrounds, and access to cutting edge technology, we had no doubt about our ability to build this first-of-its-kind robot that would transform the agriculture industry. It was going to be a mad success, and we were going to be famous!

Quadric.io Co-founder and CTO Nigel Drego, working on a vineyard robot prototype, 2016

We had just one problem: we couldn’t make it work. I mean we made it work — we wrote code and developed software and designed and built hardware prototypes that functioned (see picture). But we soon came to realize these prototypes would never become the affordable, scaleable, commercially viable products we had envisioned.

Why not? In short, the existing technology platforms simply didn’t support the autonomous functionality we were seeking. They required so much power that by the time we loaded on enough batteries to get through a few hours work, they became too large and unwieldy to move up and down vineyard rows. Equally as problematic was their slow processing times. The time from sense (taking in environmental data) to control (performing actions based on data algorithms) was impractically long.

This is where our story takes a turn. The technology didn’t exist to do what we wanted to do. So we figured we had to make it ourselves. We needed faster, lower powered processing than had ever been built before. Intel’s CPU’s worked very well for standard computer processing. And Nvidia’s GPU’s ushered in astounding new graphics processing for gaming and much more. But what we needed was something specifically designed for processing on the edge, something that supported realtime autonomous decision-making. So that’s what we decided to build at Quadric.

The promise of the autonomous future has long been touted. But other than early, limited self-driving capabilities and self-propelled vacuums that can whir around and pick up dog hair, we’re still waiting. Where are the innovations that are going to make life easier? To give us more time to do the things we love? The answer is still “it’s coming,” and the reason it’s taking so long is because of the limitations of the underlying processing technology. We believe this is a tremendous market opportunity, and we’re on the way to delivering a completely rearchitected supercomputer for edge processing.

Quadric Supercomputer (Autonomy Application)

Approximate size: 4” x 6”

Many companies have attacked AI and machine learning, developing neural networks that represent big steps forward. But we think we’re the only company that has looked at the needs of the entire end-to-end system to develop a new full-stack solution for sense and control on edge devices. Tesla recently announced its own proprietary chips because other commercially available options weren’t cutting it. This illustrates the problem we discovered. But for companies who don’t have the resources to build this tech from the ground up — or for industries that haven’t seen as much innovation as transportation yet — Quadric is a plug and play option that eliminates the need for building heterogeneous systems with significant hardware and software integration costs — thereby taking years off of product development roadmaps.

Our solution includes:

  • Plug and play support for existing sensors
  • Pre-optimized software libraries in a variety of domains such as Computer Vision, AI, Sensor Fusion & Real time control
  • Choice of power levels ranging from 250 milli Watt to 20 Watt
  • Early testing has shown results up to 100X faster and 90% reduction in power.

Read more in our whitepaper here.

So — turns out our story will definitely have a different ending. We hope that what we’re doing will not only enable the robot we envisioned to transform agriculture, but will enable real transformation of other industries as well, helping to realize the promise of an autonomous future….where dirty or difficult work is done by technology, enabling humans to have higher quality jobs, and more time to do the things we enjoy.