It took exactly two minutes for today’s TC Sessions: Robotics fulfillment panel to make its first Amazon listing. The retail giant looms above the category like no other. It played a fundamental role in the 2012 acquisition of Kiva Systems that spawned Amazon Robotics, and it remains the 800-pound gorilla that looms in the background of every conversation about warehouse automation.
Over the past ten years, the company has shown impressive dominance. It has helped the company establish a once-impossible standard of next-day—and even same-day—delivery for many orders. Retailers large and small have looked for ways to stay competitive and fuel the growth of an entire industry of warehouse robotics companies such as Locus, Fetch and Berkshire Gray.
Kiva’s 10-year acquisition remains the foundation of Amazon’s game. The mobile systems are in fact the ground floor for a modular ecosystem.
“When we look at our technology and architecture, we look at each sub-component and the capabilities that open up to us,” Amazon Global Robotics VP Joseph Quinlivan told me in a panel at the event today. “And how can we commoditize that and – like the software space – build a clean API and a form factor around it, which can be reused in many different robotics solutions and architectures. One of the reasons we were able to go pretty fast and have a We were able to build a wide range of products beyond the original Kiva product is that we were able to use this architecture and these technologies that were more than just a problem that we solve immediately.”
At Re:Mars a few weeks ago, the company unveiled a number of new robotic systems designed to fit into that growing warehouse ecosystem. The headliner of the bunch was Proteus, a new system that introduces full autonomy while retaining Kiva’s raw form factor. The company then noted:
Proteus autonomously moves through our facilities using advanced Amazon-developed security, perception and navigation technology. The robot is built to be automatically directed to do its job and move around employees – meaning it doesn’t need to be confined to restricted areas. It can operate in a way that improves the simple, safe interaction between technology and people – opening up a wider range of potential applications to help our employees – such as lifting and moving GoCarts, the non-automated wheeled conveyors used to move packages through our facilities.
We suggested at the time that the system could be the product of Amazon’s 2019 acquisition of its Boulder-based autonomous carmaker, Canvas. However, Quinlivan says the robot was developed independently of the acquisition.
“That was developed internally by the Amazon Robotics team that grew out of the Kiva acquisition,” the executive says. “A lot of times at Amazon we have concurrent development efforts. We’re excited about what the Canvas team is going to deliver and they’re going to focus on another application that we haven’t announced yet.”
He also rejects the idea that Amazon’s recently announced $1 billion fund, which backed a number of robotics companies, including Agility, is a pipeline for future acquisitions.
“I don’t think we invest — especially in early stage companies — because we want to acquire it. We hardly ever have that discussion. We invest because we believe people are passionate about what they solve, it’s an interesting problem. We almost have the mindset that we want to invest in things that we don’t think can be achieved because we could be wrong.”