Macitynet met Colin Angle, iRobot's deus Ex Machina, the company behind the Roomba phenomenon that actually brought the first real mobile robot to the homes of millions of consumers.
Whether it was an automatic vacuum cleaner and not another non-random tool or aid: the project was born from a long series of studies that resulted from years of research at the Massachussets Institute of Technology but even before the passion we have to say literally innate of Colin Angle for Robotics.
The meeting took place both with a series of questions from colleague Mark Perna in which we actively participated both during the joint session and in a direct interview with Colin.
How did the passion for robots in Colin Angle come about?
I have always cultivated a passion for discovering how things worked: I can tell you that at three years of age at home we had a problem in the toilet: I asked my mother to read me the manual of the repair instructions and under her guidance I managed it without problems. At 10 I studied a robotic crane to move the dishes in the kitchen as if it were the most natural and necessary thing in the world.
During my years at MIT my initial goal was to work on a project as cool as possible or at least that was my intention actually I think I have thought about robots and worked on projects with robots for my whole life.
If I had not done this I think I would have become a kind of forest ranger: I am a camping enthusiast.
After all, the activity that I do outside of iRobot has led me to create a non-profit organization that deals with robotics in support of nature.
The first Roomba was presented in 2002: how was what was in fact a pioneering project accepted at the time?
It was actually implied that robots would come into our lives but in 1990, when I co-founded the company with my MIT professor at the time, the aim was not to create a specific robot but to turn the robot and we explored 20 different business opportunities to create a device that was not only cool and salable but also generated a profit: in short, it could transform itself into a sustainable business.
Today we see in fact that the traditional dead vacuum cleaner or at least its market is shrinking more and more thanks to the prevalence of robotic vacuum cleaners on the one hand and manual systems (Dyson type wireless vacuum cleaners).
I am in fact very proud of what I have achieved with Roomba and after 20 years of study of preparation and presence on the market I think we are still at the beginning of this experience with still a lot of room to grow.
The strange fact is that it is very complicated to conceive and build these objects, much more than what people can believe.
What are the mistakes of the past that have prevented or prevent a greater spread of robots in our homes?
The challenge of these objects that the consumer has different expectations and many companies have failed because they thought of building anthropomorphic devices expecting people to have paid anything to take home a robot that did everything.
At the beginning of our activity nobody would have invested a lira in the project of our robots because we didn't have a concrete plan to pay our collaborators at the end of the month (iRobot did not receive any help from Venture Capital) and therefore we asked ourselves: what kind of robot should we have produced?
A robot that created more value for its cost and that allowed us to carry on our business: in the end we came to the choice of creating something that was cool but also useful and this actually creates value: the fact that the our first robot was a vacuum cleaner given by its function of performing tasks that people find boring and repetitive: the value of the invention was to educate the population to understand that the robot could do something for us.
When I started my adventure the real challenge was not so much that the robot worked but that the public realized that after decades of promises that was the time that robots entered everyone's life.
To date, robots are not yet a commodity, what can we expect from the improvement of these devices?
For 27-29 years I have built and designed robots and my main mission has been to make them become much more autonomous: the perfect robot the ones you don't see and don't touch. Today we are working a lot on artificial intelligence but much more intelligence than what makes a robot autonomous: useful intelligence comes from collaboration.What can be obtained by interacting with humans or those who have greater knowledge of the environment.
For example, if we send an astronaut to explore an unknown planet, he will be able to make autonomous decisions based on what he sees but if we have a camera on his shoulder connected to a coordination center, he can be helped to identify things that at first glance escaped him. or to inquire with data that are unknown to him personally or to advise him to perform a certain action.
The other aspect, in addition to the combination with the interaction with a control unit, is that of being part of a system that provides, always with the example of the astronaut, the rocket, the landing module, the rover for the surface and the return module to the mother ship. : everything must be planned and combined perfectly.
If you ask me what is coming in robotics I can answer that it will improve the quality of team work with other devices and with the end user.
You mentioned AI, for humans logical thinking is a natural attitude but for the robot something that you have to insert in its capabilities. Robots can learn from home automatically
The current problem with how intelligent robots can be, not AI, but an understanding of the surrounding environment. Let's think about one thing: why do commercial robots have no arms? Because they don't know what they are up against. An ideal situation would be to have a robot I order: go to the kitchen and bring me a beer. The problem may not be so much to make you understand where the kitchen (the Roomba Robots already have a house plan and can identify the various areas) but how to identify the position of the beer, discriminate the color of the can, the brand etc: in fact we must have more information about the house and what is inside.
Since we are talking about artificial intelligence: if you were to give advice to those involved in self-driving cars what would you suggest?
They are very different areas: if a robot hits the foot of a person who has no problems at home, quite different if a car hits your leg. In any case, I think that to have real generalized autonomous driving applications twenty years will pass while having vehicles such as buses and means of transport on a determined and controlled route is a much more current event: we have created an expectation in this sense and we must give an answer .
What is the relationship between your Roomba and the smart home?
Let's say that the smart home as conceived today has not reached a real level of utility: not yet up to its potential.With Roomba we are creating a system that identifies the house and learns to know it but the request for information is still very high.
Could it be possible to manage everything with cheap tags? The dog bowl that is continuously moved, the bottles with a description of the contents.
I think the solution is a combination of what the robot can see and the definition we offer it: this collaboration between recognition intelligence and user help would allow to obtain results that allow to identify objects to satisfy a request, to avoid a collision fortuitous or an unwanted shift.
And here we come to another question: how are customer data managed: the acquisition of the rooms in the house, the presence of objects, the plan of the house itself and the relationship between the spaces?
In iRobot we want to be a DATA 2.0 company for how we manage user data. Normally when users notice that for some reason their compromised privacy gets very angry, they protest vehemently but then the next day they buy the cheapest device on the market.
Security and privacy come at a cost and on the one hand the pact between company and user must be clear to understand what remains in the device and what can optionally go on the web and on the other hand it must be understood that keeping data in a private environment has a cost that is reflected in the cost of the product and its reliability.
What I hope that a third-party organization capable of certifying the quality of the product from the point of view of privacy and safety and reliability, a sort of quality mark like the one I find on the milk bottle at the supermarket (Colin Angle he lives in Bedford, MA) which certifies me the production and the product chain in all its aspects.
I am working hard to get to a certification body that provides for the analysis of each of the aspects of the supply chain in the iOT and consumer electronics field: from the choice of the microprocessor to the data encryption system to the cloud storage system of the same.
Among our long-term missions is to create products that are efficient for a long time and to make the place where you raise your children comfortable and healthy. At the same time we must think that the care relationship towards the elderly has changed: the greater longevity leads to reduced families, single life in single-family households and this corresponds to the need for help that can be given by robots like iRoomba: we can also contribute with this to reverse the trend of a lifestyle of the elderly that worsened compared to previous generations. As we said, iRoomba was created to do a job that users do not want to do in first person and this is even more true for those who find it even more difficult to perform cleaning operations.
For us the solution is to use robots to obtain a better quality of lifeand have more time to do what we really like to do.
Why should I buy an iRobot product today since every year you come out with a novelty that makes the previous product obsolete?
In reality, the robots produced in recent years are no longer products for themselves but platforms that we can update with new functions as happened recently to make them even more efficient and proactive for their tasks. right now we are thinking about updates every 3 months and then slow down a little further.
What can you tell us curious about how Roomba are used in the home?
Well the most curious thing that users tend to gain confidence after about two weeks naming the robots that go around the house_ in the end they consider them as a pet and despair when for some reason they fail to finish their usual tour.
There are also studies in Japan on how house organization changes before and after the arrival of a cleaning robot.
It can also be said that robots are an educational tool: for example, they can teach kids to keep their rooms tidier if they want the robot to work unimpeded
Absolutely yes also here we discussed the fact that a robot somehow breaks the tension between generations: the boy arranges his things around responding to a command from the parent but because in fact if he does not prepare the environment correctly this will not be cleaned at must.
Even if you can't tell us about the ongoing projects, what can you tell us about the products you plan to make in the coming years?
To make a consumer robot you have to work on proven and easily achievable technologies: for this reason I think that we will hardly have robots with legs in the immediate future while it might be easier to get robots with arms: I basically define myself as a practical robot Guy and I'd like to make a robot with arms capable of moving freely: a sort of segway with arms that identifies objects and can accompany humans in their daily actions, for example unloading or loading dishes in the dishwasher!
We then spoke with Colin Angle about the next iRobot product that will hit the market: a lawnmower with innovative features that has a particular link with our country. But we will talk about this in another article.