QUESTIONS. AND ANSWERS
So let’s get into the Q & As.
Q: Tell us about BUBS
A: BUBS is a second-generation delivery robot. It is a large unit relative to what we typically would see out in the market. It uses a suite of sensors onboard to give it spacial and environmental awareness. We use systems like the NVIDIA Nano to process data that the robot sees in real time, which allows it to make its own autonomous decisions to navigate through the world.
The robot is large-capacity and has a locking, remotely operated lid that allows it to securely hold and transport whatever it is our customers are interested in moving. It’s a large-enough system that it can hold food, beverages, product from a store, pharmaceuticals or lab samples from a hospital, dirty linens in a long-term care facility, product on a golf course – really anything that needs to be transported from Point A to Point B.
It has indoor and outdoor capabilities, so it is weather-resistant. It can handle some pretty significant slopes and terrain. The robot, in addition to its autonomy and sensor package, utilises a radio system that allows us to have remote assistance or piloting as necessary through a WiFI connection, a 4G/5G connection, or private network. We currently use the Rogers 5G network as our backbone, and from a teleoperation or semi-autonomous perspective, we can have an operator located anywhere on planet earth and have that person remotely assist or operate the robot as necessary with under a 1/10th of a second latency. So it’s near real-time using EDGE computing.
IS BUBS DESIGNED FOR AUTONOMOUS OPS OR TELE-OPERATION?
That was the question. Here’s the answer.
A: The answer depends on the application itself. So some environments where we have a high degree of predictability and can pre-map and understand that environment, those are environments that are more conducive to fully autonomous operations. The robot can be trained using Computer Vision and AI to autonomously navigate through an entire space if that space is predictable.
Alternatively, if there is a high degree of variability, or there are safety or regulatory concerns that require a human in the loop, we have that option as well. So, for example, if a robot was to be traveling in a city environment and it needed to cross a road – that’s a complex procedure for any robot to do. And there’s likely a degree of human interface that would be beneficial to have that robot determine when and where it’s safe to cross the road. Or if a path was blocked by a large-enough obstacle, and the robot needed to exit a geofence that is pre-programmed into that operation, in order to safely manoeuevre around that obstacle, it’s likely a good idea to have a human in the loop to make that complex decision.
The more repetitive times that type of an application happens, the more a robot can be trained to autonomously execute those types of scenarios. As that robot’s deployment increases over time, the human interface required decreases. But there will always be some level of human in the loop.
WHAT ABOUT REGULATORY ENVIRONMENTS AND ADOPTION?
A: Over the next 10 years, we will see an enormous increase in the reliance on robotics to do basic things like delivery inside municipal environments. One of the things I learned at Aeryon years ago was the importance of engaging with government early on, because government can otherwise potentially shut down your operations at a really inopportune time. And so Real Life Robotics has already engaged with a number of Canadian cities and had early approvals to allow our robots to drive around in certain automation projects in city environments.
Municipalities typically have concerns around full automation and Level 4, Level 5 autonomy. If Elon Musk and his team are not able to get approvals to drive around in downtown Toronto, how do we think we’re going to get the same approvals to drive around autonomously? We’re not. So the cities have actually really embraced the fact that our robots can have a human in the loop to make some of those difficult decisions. That helps alleviate some of the concerns around full autonomy. But we have spent the time building the groundwork to allow us to operate in their environments and we, in return, intend to work very closely with those cities to actually build the playbook, and build the ruleset and the framework around successful and safe deployment of robots in urban environments.
CAN BUBS BE MODIFIED FOR CLIENTS?
A: Absolutely. Our mandate is to commercialise robots. And as part of any startup growth plan, sometimes there are pivots along the way that you need to make. But in general, a client that has a real ROI potential where robots can facilitate that, and a client that has the potential for scale, that’s our expertise. With the combined benefit of having InDro, we can not only develop a very specific robot solution to solve a customer’s immediate concerns or challenges, we can also scale that robot.
Q: Why did you feel InDro was the best fit for a partner?
A: In general, InDro would be considered a world-class R&D company – hands down, bar none. And that’s why we partnered with them. The firm has an enormous skillset, including expertise with autonomy, sensor fusion and integration. Because the company has all off that, plus a large engineering staff, we’re lucky to call InDro a true partner. InDro’s capabilities and agility will help speed the path of Real Life – and our clients – to commercialisation.
Below: BUBS in action during a pilot project in Surrey, BC