(MENAFN- Robotics & automation News) Interview: Ambi Robotics founder sees 'huge opportunities' in new markets
December 28, 2024 by David Edwards
The founding of Ambi Robotics is somewhat obscured by the mist of history, but it started with some abandoned robots in a basement and the co-founders meeting at University of California Berkeley in 2018.
Within about a year, the five founders made a breakthrough in simulation-to-reality transfer learning systems for robotics, calling it Sim2Real AI, which, in turn, led to the establishment of Ambidextrous Laboratories Inc, and a year later to the commercialization of the technology in the form of the AmbiSort parcel sorting system.
The lightning quick pace of development led to huge interest from investors, who poured in $21 million into the startup by 2021, and another $32 million the following year.
Now, three years on, Ambi Robotics can look back on numerous multimillion-dollar commercial installations and look forward to many more diverse deployments in the years ahead.
In this interview, we speak to Jeff Mahler (main image), one of the founders of Ambi Robotics and its chief technology offer, and ask him about himself, his company and the future of robotics.
Robotics & Automation News: Let's start by telling readers about yourself and your job at Ambi Robotics.
Jeff Mahler : I'm Jeff Mahler, co-founder and chief technology officer of Ambi Robotics. I lead the engineering and product teams at Ambi, executing our technology strategy to develop highly dexterous AI-powered robotic systems that help people handle more in commercial logistics operations.
I earned my Ph.D. from UC Berkeley, where my research focused on AI for robotic grasping. It was in Professor Ken Goldberg's AUTOlab where I met Stephen McKinley, David Gealy and Matt Matl, and we began combining our complementary skills to later become the co-founders of Ambi Robotics.
My experiences in both industry and academia, including co-launching UC Berkeley's course on robotic manipulation and publishing award-nominated research on robotics and AI, have given me a strong foundation to develop advanced technologies and lead the creation of transformative robotic systems at Ambi Robotics.
R&AN: And let's also introduce the company, Ambi Robotics, to readers. What are the main services and products and what is its position in the market.
JM : Ambi Robotics is an AI and robotics company developing advanced solutions that scale ecommerce operations to meet demand while empowering humans to handle more.
Our flagship solution, AmbiSort A-Series, uses robotic arms combined with advanced gantry technology to automate the sorting of mixed parcels into destination containers.
It's modular and configurable, supporting customer demand for both bin-fed and conveyor-fed operations for high-efficiency parcel sorting.
All of our robots run on our proprietary operating system, AmbiOS, which uses simulation-to-reality (Sim2Real) AI to train the robots 10,000 times faster in virtual simulations and then transfer those learnings to the real world, enabling the robots to handle any bag, box, envelope or unpackaged item with high speed and accuracy.
With continuous learning capabilities, our technology's performance improves and evolves over time.
Ambi Robotics is leading the way for AI-powered automation to solve operational challenges for shippers, positioning us at the forefront of innovation in the booming e-commerce shipping and logistics space.
R&AN: What are the key problems that Ambi Robotics solves? When Ambi Robotics is asked to do a job, does it take over from human-centred systems or does it generally improve on a previously implemented technological system? Maybe give us a use case or two.
JM : We're tackling some big challenges in ecommerce and supply chain, especially with the increasing demand for faster and more accurate parcel sorting.
The AmbiSort A-Series system automates the sorting of mixed parcels from deep bins or automated conveyor induction into destination sacks, a traditionally manual and time-consuming process prone to errors, injuries and bottlenecks.
By bringing in our AI-powered robotic systems, warehouses can boost throughput, accuracy and cut costs-all without needing to completely overhaul their existing process. Powered by Sim2Real AI, solutions from Ambi Robotics are ready to go from day one and add immediate value and cost savings.
For example, during a partnership with a global shipping and mailing company, we deployed our AmbiSort products into their existing middle-mile operations to help automate and speed up parcel sortation.
During our time together, we sorted 87 million packages and trained over 400 of their parcel handlers for higher-value, certified robot operator roles.
With this move, their employees were able to focus on more complex and higher-value work while cutting operational costs.
R&AN: What are the current challenges in parcel sortation? And how do you see the future playing out in terms of parcel sortation technologies?
JM : Many warehouses run into problems with labor shortages combined with rising parcel volumes, especially around the peak holiday season, and inefficiencies in their traditional technology that may cause delays and errors.
They also often struggle to scale operations to meet growing ecommerce demand while keeping costs manageable. For example, rising transportation costs are an ongoing challenge to meeting customer desires for lower shipping rates.
Looking ahead, we see advanced robotics and automation taking on a larger role within warehouses, leveraging AI to create highly repeatable, efficient, scalable and flexible systems that can integrate with human workers and existing infrastructure to handle diverse parcel profiles efficiently.
In addition, intelligent automation can create a fountain of data on every item handled in order to empower operations management to make smarter decisions on how to boost productivity.
R&AN: At the moment, most of the end effectors on robots tend to be suction devices of some kind. Is there an alternative to this? Do you use any soft end effectors or human-like hands to deal with specific kinds of items? Is this diversification something that is relevant for you?
JM : We primarily use suction-based end effectors because they are highly reliable and effective for handling a variety of parcel types, especially ones with non-porous, flat surfaces like cardboard.
Our robots have multiple suction cups of different sizes and flexibility, so they can choose the best“fingertip” for whatever they're handling. While suction cups work great for a lot of items, we know there are objects, like irregularly shaped or delicate ones, that might need a different touch.
To address this, we're looking into other types of end effectors, like parallel grippers, to extend the range of items our robots can handle. This flexibility is key as we keep improving our robots to handle an even wider range of objects efficiently and accurately.
R&AN: What is the overall market potential for a company like yours? How big could it become, given the growth in e-commerce and so on, and considering the competition you are already up against?
JM : The market potential is huge, especially with the massive growth in ecommerce. In 2022 there were over 150 billion packages shipped worldwide, and each one of those packages was touched by a human hand about 10 times.
That means that there are over 1 trillion package“touches” occurring each year, and this number is growing rapidly. As more people shop online, the demand for fast, efficient and scalable fulfillment solutions is only going to keep rising.
With robots like ours that can handle parcel sortation and even adapt to different types of packages, we're in a strong position to meet that demand.
Our competitive edge is the ability to handle a wide variety of objects with precision and speed, thanks to our AI-driven system, and our ability to easily integrate with whatever warehouse layout is needed.
As automation becomes a bigger part of the supply chain, we see a ton of room to grow and capture a significant share of the market.
We're working alongside global leading logistics companies and continuing to evolve our technology to meet their needs as they grow.
R&AN: While you currently specialise in parcel sortation, was that always what you had in mind? And could you diversify into new markets in the future. If so, which markets?
JM : Ambi Robotics is all about real-world impact, and parcel sortation was where we saw the biggest opportunity to apply our technology due to the massive market opportunity.
That being said, we've always had the vision to make our robots more versatile, and as we grow we can definitely see ourselves branching out into other markets.
There are huge opportunities to provide robots that automate repetitive motions in other industries with high-volume material handling such as ecommerce order fulfillment, retail distribution, manufacturing and even recycling.
MENAFN28122024005532012229ID1109037657