Stephen Delaporte of Avar Robotics Answers Four Questions for Founders
Avar Robotics is developing industrial robots to provide a lower cost and more flexible alternative to large scale industrial robotic arms. Avar’s core launch product is a universal robotic platform that can be adapted for different industrial applications through different custom-built end effectors but is initially focused on material handling.
1. Why did you start Avar Robotics? Do you remember the moment when you first thought of the idea?
Several years ago, I had become intrigued by the evolution of what is sometimes referred to as high-density robotics, where many robots within a controlled environment can work together to perform tasks such as moving material around within an order fulfillment center. My initial interest in this area of robotics actually stemmed from another robot I was developing back in 2012 which was essentially an early version of what evolved in the marketplace as virtual assistant devices, like the various Amazon Echo and Google Assistant AI products. At that time, I had become friendly with the founders of Kiva Systems after demoing my consumer robot for them at Columbia University, and they were intrigued enough by what I had built that they began to regularly advise me and my startup. They were also kind enough to invite me to see a live demo of their system at their facility in North Reading, Massachusetts. Kiva Systems, which created the robots that are used within Amazon’s fulfillment centers, really pioneered the idea of using mobile robots within these environments to provide a more productive solution over the pre-existing methods used, so it was an eye-opening moment for me to see their assembly line and the inner-workings of their technology in person. One of the hardware components within their mobile robot assembly that was most intriguing to me was their vertical lifting mechanism that could pick up a shelf with hundreds of pounds of weight. The engineering behind it was extraordinary! However, after learning more about how their robots worked to bring shelves filled with items to human pick workers, I immediately asked myself if this was the most efficient solution, and also why the pick worker was the only component to their system that wasn’t automated, and from there it became an ongoing problem that I thought about.
Fast forward a couple of years later in 2015, and I had really started to immerse myself in trying to figure out if it was possible to automate the operation of a pick worker with a simple solution. I had also started to investigate the tasks of a pick worker by actually speaking directly to people who were employed at the Amazon fulfillment centers, and I soon learned that the pick worker operation could not only take its toll on the body, especially because of the constant kneeling and moving around of items with such a high demand for pick rate speed, but that generally people didn’t like to do a job where there was an intense amount of repetitive labor throughout the entire workday.
In 2016 I started to build out a robotic solution to this problem by integrating a gantry configuration with a carrier plate and a unique gripper system that used a four-bar linkage that could grab an item and then eject it with only one actuator and a few sensors. Typically, a gripper will use many actuators (or motors) and many sensors just to grab an item and then release it, so my solution was simple enough that the entire robot could be made at a lower cost than a standard solution such as a robotic arm with a gripper that integrated many motors, moving parts, and sensors. I developed and built the whole robot part-time over a 5-6 month period, and then from there, this attracted a small team of engineers and people with executive level management experience. It also helped us attract the attention of some of the largest companies in the world that we then began to collaborate with.
Once Coda was built and proven as an alternative approach to using a large-scale robotic arm for sorting items, it then became clear that the next iteration of our robot could use a similar kind of gantry system and gripper, but with wheels integrated so that it would not have to be constrained to a belt drive and predefined structure, and this idea evolved into our Prefius mobile robot, which is what we’re entirely focused on developing now as a company. Prefius is unique in that it can grab a large variety of items with different sizes, weights, and shapes, and then store them locally on the robot before it brings those items to an endpoint. The Prefius mobile robot really combines the world of mobile robots with the world of end effectors and grippers so it can fulfill an entire order or multiple orders in one journey. The typical mobile robotic configuration today uses a robot that can move an entire shelving system from point A to point B where a pick work then picks out a single item, and that same robot then brings the shelf back to its original position on the warehouse floor or to an entirely new position. The inefficiencies that exist with this approach include the fact that the shelving system that the robot is carrying around is made up of mostly items that represent dead weight. Now, if that same robot could pick up several items along the way to its final destination like a bus would, it would end up with a much more efficient result and with greater accuracy as it becomes completely autonomous. Our approach with Prefius is advantageous for not just fulfillment centers, but also for stores of the future that will integrate automation to pick orders at a faster rate with greater efficiency for consumers.
2. How do you see Avar Robotics impacting how we live our lives?
Our mission as a company is to create the world’s most cost-effective and productive automation and robotic technologies, and by achieving this, we believe that people will see an increase in their work output and also be freed from labor-intensive jobs that are undesirable while new kinds of jobs will emerge that can nurture their minds more, such as monitoring the robots remotely or analyzing data that has been collected by them, along with maintaining them, and so on. We as humans are one of the most adaptable creatures on the planet — children who have had half their brain removed from a hemispherectomy to eliminate severe seizures, can regain cognitive functionality that would normally be performed by the missing hemisphere within the hemisphere that remains intact. This is fascinating and shows how truly adaptable we are. We can adapt physically, emotionally, and intellectually, and ultimately we can adapt and be retrained for new kinds of jobs as new technologies continue to redefine what a job is and how it can best be carried out.
3. How have New York and Newlab positively impacted the development of your business?
It has been a lifelong dream of mine to build a successful robotics company, and I’ve been naive enough to work towards making this dream into a reality, but over the years, after attempting to start robotics companies on a couple of occasions with little success, I learned the hard way about how difficult it is to build a company implementing its own proprietary robotic technology. I’ve been fortunate enough to have gained a lot of traction now with Avar Robotics, and I would argue that a lot of this recent success can be attributed to Newlab’s resources and community and the growing NY tech hardware startup ecosystem. I can literally walk down the main space of Newlab on any day of the week and see a dozen different startups at work, and simply walk up to anyone and inquire about what they are learning, how they have approached funding and the development of their unique technology, or even ask about an esoteric technical problem, or some alternative way to fabricate something, and collectively we will often come up with new solutions or gain new knowledge from these daily interactions. It’s an amazing place to work every day. Everyone really feeds off of each other’s entrepreneurial and inventive energy, and when you’re doing something as difficult as building a hardware startup, moral and intellectual support from fellow founders is enormously invaluable.
4. What do you see as your responsibility to make technology ethical and accessible to all?
There is often a misconception that as robotic technology becomes more pervasive, it will eliminate jobs and be disruptive in a negative way, but I would argue that this perspective is mostly incorrect in that the vast majority of new technologies that upend traditional ways of doing things, not just robotics, simply augment work output, and remove the inefficiencies of tasks and job operations which ultimately can provide new kinds of jobs.
When VisiCalc, for example, was first launched on the Apple II computer in the late 70s, it was the first spreadsheet software program and subsequently became the “killer app” for the Apple computer. What the electronic spreadsheet did at its core though, was to act as a multiplier for work output in that it could calculate budgets where one change to a single defined calculation and number could impact every other calculation and number, so the work that many people would do in an hour, could now be done by one person in the same amount of time.
This is a great analogy for what robotics is doing today, in that it is simply acting as a multiplier for work output. The only difference is, in the case of the robot, the labor is physical, and the notion of it taking over a simple job such as moving an item from point A to point B, for example, tends to be much more troubling to people because it is less abstract than the idea of an electronic spreadsheet more efficiently doing calculations, but in the end, they are both eliminating inefficiencies in work output, and thus re-inventing the role of a job as it relates to its particular task. I think in this respect, robots represent a major paradigm shift for our future regarding how we function as humans, and because of this, they need to be safe and collaborative, or carefully isolated if there is a major safety concern when they are used in industrial environments, all while still being user-friendly to interact with when controlled and monitored remotely so that people, in general, will start to feel more comfortable with the existence of our fellow robotic machines as they continue to become more ubiquitous.
Want to join over 150 member companies and 800 entrepreneurs, engineers, and inventors tackling the world’s biggest challenges?
Apply for Newlab membership.