Author Ahti Heinla, co-founder and CTO at Starship Technologies
I see robots every day. I see them walking down the street at the speed of pedestrians, standing up to make sure it’s okay to cross the street. Sometimes I find them talking to pedestrians. I am a bit of a viewer of the technical concept – AI Wondland. But this is not a fantasy, no dream, it is a reality that our dedicated vision team has built over the past 5 years; we have brought the future into the future.
As recently as a few years ago, these robots needed human help and were included in their journeys, much like the independent car manufacturers, who test their vehicles in public using ‘safety drivers’.
Starship became the first group of robots to start operating regularly in public places about 18 months ago, without the use of security operators; we allow our robots to explore the world for themselves. We now use our robots every day in several cities around the world, bringing people their food, packages, and food.
Sharing knowledge and gaining knowledge
It’s fun to be the first.
Back when I was a startup engineer in Skype, we were the first to make Voice over IP available in a practical way; we are also working to do the same with robots in public places. For four years, our engineering teams have been working secretly to achieve what has been successful and amazing.
I want to share with you some of our creative journey. In the coming weeks and months, other members of the Starship expert team will share their tour.
During this trip we have been working with computer vision, designing solutions and identifying obstacles – well-researched studies in the field of robotic learning. True, Starship started out as a research project, but it soon turned into a supportive activity.
This means that in addition to the well-planned Levenberg-Marquardt algorithm for optimal performance, we need to develop programs that:
- Save most of our sensors – after all, we don’t want to spend more hours with hands; we have made hundreds of robots and are preparing for a major mission.
- Think about the amount of energy that can be generated each time from a robot battery – so that we can design robots to send, depending on the condition of their batteries.
- Predict how many minutes it will take for restaurants to prepare food – so the robot just arrives on time!
Many of the independent robots that exist in the world today are expensive, built as technology or search engines, and are not used for commercial purposes. A package with a stand-alone tool can cost $ 10,000. This does not work in the supply chain, it is not a high-end business for which you can pay a lot of money.
Self-propelled search engines usually have three kilowatts of power per machine; impossible for a small safe robot. As a result, part of our engineering journey has been to create very little economy. Here are the topics to consider:
- Image editing on interactive reading platforms.
- Working around the complexities of software development.
- Track how many robots need to be repaired, and why.
- Creating state-of-the-art machinery, ensuring that we use our machinery properly.
The tour was also of a clear design, combining a number of paintings, drawings and explorations before making our first plastic body.
Going back to the early days when we were still smart, we didn’t want to describe what our robots look like. Frequent public testing requires the use of a garbage bag, engraved on the body of a robot as a hiding place!
Designing working robots is a way to match science, architecture and cost. The integration of the various classes is a Starship tradition. There is nothing easy in a robot. What you know about what happened is possible; all sensors have faulty and faulty methods, and even a task that seems as simple as designing a robot to stop obstacles it could be his little research program.
Starship is a fast-growing business and it is important that you do more than just research. Professionals who enjoy Starship are often not real scientists, not destroyers, not real experts; they have some of these qualities and can use them in connection with the work they are doing. We need complex technical solutions to be implemented quickly and on the challenges of low-cost weapons.
Being wise and careful is a valuable skill.
The Sabbath is a long time in Starship
Earlier in the week, our team will use a new method to detect cloud movements and return them to the test list overnight, testing them on our secret tests over the weekend.
It will be on the streets next Monday, the group has already reported on its attendance at the Monday Monday meeting. Last Monday some of the engineering team said they were earning a 300% + profit on one of their achievements, last week.
More as a result and a growth guide
Metric and data are a major component of Starship technology.
You see, back when we were just starting out we had no idea – we hadn’t really traveled much yet. Every day we change our robot (yeah, one time that), take it to the trail and see how it goes. Now we have many, driving around every day – too many for the engineers to see directly.
Thanks to our discoveries, we can now see how our robots work, hundreds of them. We can do weekly data dive seminars, where engineers share their findings and view donations randomly to connect with their work.
As we work to make our robots run smoothly, we analyze the content of our ‘Fast’ table; there are at least 1 billion lines on the table. Some of the tables include ‘cross-road events’, our maps, any command every robot has received from our servers, and obviously data collected from each arrival they make.
Four years ago, we didn’t have all of that. Back when we were just starting out – and we don’t run a business right now – I often had to convince people that robot shipping really worked. People find it hard to believe and are quick to point out various reasons.
Do doubts and fears always accompany new technology?
A few years ago, I arrived at New York Airport in JFK with a robot in my bag. The traditional man asked: “What is this?” I explained that it was a roadblock, and they replied: “Friend, this is New York! It will be stolen within minutes! ”
In fact, at that time almost everyone thought these robots would be stolen – I hope they would (postal votes were stolen, even if they were rare). So far our robots have traveled 200,000km (130,000 miles) and we have never seen this problem.
There are protections available. The machine has a siren and 10 cameras, is always connected to the internet and knows its location accurately by 2cm (thanks to the above-mentioned Levenberg-Marquardt algorithm, with 66,000 lines set up by the only C ++ that enables our robots to operate).
People think that pedestrians may be afraid of robots on the road or may not accept their presence. Will people call the police? In fact, we could not even be sure of that! However, when we lost one of the robots outside, we were greatly surprised.
What happened next shocked us: people just ignored it. Most people didn’t care about robots, even the first ones, and people weren’t scared. Some take out their phones and post on Instagram of how they have seen the future.
And that’s what we want.
We want people to take care of our robots just as much as they do the dishwashers. The process of quietly receiving robots as if they were always with us has been repeated in every city around the world where we have been working.
It’s all right. As soon as people hear that these robots are really helpful to their neighbors, they are drawn to one another. Children also write letters of appreciation for the robots, we have a ‘thank you wall’ to confirm this!
Driving the final milestone would not be easy, and we knew it would be a daunting task. We always knew that there would be a number of obstacles that needed to be overcome – there would be hundreds of obstacles! But we realized long ago that these problems are over – they just need skill and perseverance.
Some startups start like running sprint, and throwing Together Minority Products within three months. For Starship it’s like a race – a lot of unnecessary effort is required, but the result brings the biggest good in the world.
Sending the last mile is one of the few companies in the world that has not suffered much since the car was invented. The team at Starship wants to change this, and with more than 20,000 donations under us, we are on a journey.
For more information, see our second post on Technical Engineering at Neural Networks and how to design our robots here – https://medium.com/starshiptechnologies/how-neural-networks-power-robots-at-starship-3262cd317ec0