By Ahati Henla, co-founder and CTO of Starship Technologies
I see robots every day. I see their pedestrians speeding down the sidewalk, stopping to make sure the road is safe to cross. Sometimes I even catch them talking to pedestrians. This is a glimpse of the imagination of the technical mind – an AI wonderland. But it’s not a hallucination, it’s not a dream, it’s a reality that our dedicated team of visionaries has built up over the last 5 years; We have now come to the future.
As recently as a few years ago, these robots needed some human support and were with them on their journeys, followed by many autonomous car manufacturers, who test their cars in public using ‘safety drivers’.
Starship became the first robotics team to start working regularly in public spaces about 18 months ago, without the use of safety drivers; We let our robots explore the earth on their own. We now operate our network of robots in different cities around the world every day, bringing people their dinner, parcels and groceries.
Sharing knowledge is acquired knowledge
Being the first is exciting.
When I was a founding engineer on Skype, we first made voice over IP accessible in a practical way; We are now working to do the same with robots in public spaces. Over four years, our engineering teams have worked behind closed doors with a significant breakthrough and an amazing experience.
I would like to share with you some details of our technological journey. In the coming weeks and months, other members of the Starship Engineering team will also share aspects of their travels.
In this journey we have worked on computer vision, path planning and obstacle identification যে topics that have been well researched in academic robotics. Indeed, Starship began as a research project, but was soon transformed into a functional, practical delivery operation.
This means that in addition to fine-tuning the Levenberg-Markward algorithm for non-linear optimization, we also need to develop software:
- Automatically calibrate most of our sensors – after all, we don’t want to spend a few hours manually calibrating them; We have built hundreds of robots and are currently preparing for a larger scale.
- Predict how much power each trip will emit from a robot’s battery – so we can orchestrate based on the battery status of the robot we need to send.
- Estimate how many minutes it will take to prepare food in a restaurant – so the robot arrives just in time!
The autonomous robots that currently exist in the world are expensive, they are made as technology exhibitors or research vehicles and are not used for commercial purposes. A single sensor package for an autonomous device can cost up to 10,000. It won’t just work in the delivery space, it’s not a luxury industry where you can charge a premium.
Autonomous driving research vehicles often have 3 kilowatts of computing power in the trunk; Unreal for a small, secure delivery robot. Therefore, part of our engineering journey is designed for lower unit economy. Here are some things to consider:
- Advanced image processing on lower end computational platforms.
- Dealing with hardware issues in software.
- How long the robot needs maintenance, and why tracking.
- Creating an advanced route planning system so that we can ensure that we are using our robot network efficiently.
It also made quite a journey in visual design with hundreds of sketches, drawings and surveys before making our robot’s first plastic body.
In the early days when we were still in stealth mode, we didn’t want to reveal what our robots looked like. Regular public testing requires creative use of garbage bags, taped in disguise on the robot’s body!
Creating practical robotics is a combination of science, methodological engineering and hackery. This mix of different branches is the main feature of Starship. Nothing is ever easier in robotics. All your knowledge about the situation is possible; All sensors have failure modes and errors and even a seemingly simple task such as Stopping the robot Could have its own small research project.
Starship is a fast-moving startup business and it’s important not to be a big research project. The engineers who get excited about Starship are often not pure scientists, not pure hackers, not pure engineers; They have a number of these features and can be used as a handy task. We need complex technical solutions for rapid implementation within the constraints of low cost hardware resources.
Cleverness and resourcefulness are valuable skills.
A week long time at Starship
At the beginning of the week our team will implement a new algorithm to detect the carb from the point cloud and re-test it overnight against a complete test case database, they will test it live in our private test-ground week
It will be on the road next Monday, the team is already reporting on their progress at our Monday engineering meeting. Most Mondays some engineering teams are reporting a 300% + profit on at least one metric that was achieved just a week ago.
As a result data and scale are helpful
Metrics and data have become a big part of starship engineering.
You see, we didn’t have any data when we started – we haven’t run much yet. Every day we changed our robot (yes, just one then), took it to the sidewalk and saw how it works. We have a lot now, driving autonomously every day – a lot more to see the engineers live.
Thanks to the information, we can now see how our robots work, hundreds. We can organize weekly ‘data dive’ seminars, where engineers share results to communicate with their work and see random deliveries.
When we are working to make our robots run more smoothly, we analyze the data in the ‘Acceleration Events’ table in our data warehouse; There are at least 1 billion rows on that table. Other tables include ‘road crossing events’, our maps, every command each robot receives from our servers, and obviously information collected from every delivery they make.
Four years ago, we had none of that. When we were just starting out – and not yet running commercial deliveries – I often had to convince people that robotic delivery really works. People had a hard time believing and quickly showed different reasons.
Do doubts and fears always accompany new technology?
Several years ago, I landed at JFK Airport in New York with a robot in my luggage. The customs man asked explicitly: “What is this thing?” I explained that it was a pavement delivery robot, to which he replied: “Dude, this is New York! It’ll be stolen in a few minutes!”
In fact, at the time almost everyone thought these robots would be stolen – I’m sure they probably would (post delivery vans would be stolen, even rarely). To date our robots have driven 200,000 kilometers (130,000 miles) and we have yet to see that problem.
There are definitely security features out there. The robot has a siren and 10 cameras, is constantly connected to the Internet and knows its precise position with an accuracy of 2cm (thanks to the Levenberg-Marquardt algorithm mentioned above and 66,000 lines of automatically generated C ++ code that enable our robots to use it). .
People thought that pedestrians on the sidewalk might be afraid of robots or not accept their presence. Will people call the police? Honestly, we weren’t sure about that! However, once we left one of the robots on the sidewalk, we were quite surprised.
What happened next surprised us: people just ignored it. The vast majority of the public paid no attention to the robots, even those who saw it for the first time, and people were certainly not afraid. Others will take out their phones and post on Instagram how they saw the future.
And that’s what we wanted.
We want people to pay as much attention to our robots as they do to their dishwashers. This pattern of silently accepting robots so that they are always with us is what we have repeated in any city in the world.
It’s getting better. Once people know that these robots provide a useful service in the neighborhood, they build a relationship with them. Kids are even writing thank you letters to robots, we have a ‘thank you letter wall’ to prove it!
Automating last mile deliveries will never be easy, and we knew it would be a daunting project. We also knew that there would be multiple basic roadblocks that would need to be addressed – hundreds of roadblocks! But we have long since realized that all these problems are solvable – they only require ingenuity and perseverance.
Some startups start out like running a sprint, combining a minimum effective product in 3 months. It’s a lot like a marathon for Starship – it takes a lot of continuous effort, but the end result brings huge benefits to the world.
Last mile delivery is one of the industries in the world that has experienced technological disruptions since the adoption of automobiles. The Starship team is trying to change that and we are on our way with more than 20,000 deliveries under our belt.
If you’re interested in learning more, check out our second engineering blog post on Neural Networks and how they power our robots here – https://medium.com/starshiptechnologies/how-neural-networks-power-robots-at-starship -3262cd317ec0