Artificial Intelligence (AI) is changing the way we live. Both at work and at home, robotics and other forms of AI are becoming an increasingly relevant part of daily lives and routines. Robot vacuum cleaners are one of the most common forms of robotic technology used in the modern home. Almost a quarter of all vacuum cleaners sold today, are robotic. These seem like very complex machines that do amazing things. But how do robotic vacuum cleaners work?
Like all technology, robotic vacuum cleaners have evolved and continue to do so at a rapid pace. The first robot vacuum cleaners to come out in the 1990s were too stupid and too expensive to be of any use to us. Things have come a long way since then and robotic vacuum cleaners became a reality for the average consumer in 2002, with the first iRobot Roomba. Even these machines weren’t as nimble and efficient as those that we see today, but they made their mark and we’ve grown to depend on them.
The basic function of a robotic vacuum cleaner is the same as any other. It has an electric motor that creates suction and this is used to collect dirt and deposit into a dust cup. The obvious difference with a robot vacuum cleaner is that it is autonomous, meaning that it can function on its own – without human input. In order to work autonomously, a robotic vacuum cleaner needs to have some form of intelligence – it needs to make simple decisions. In order to that, it needs to collect information to make these decisions.
A robot vacuum cleaner uses sensors to read its environment. Over the years, different types of sensors have been used and modern versions use mostly laser or infrared sensors. These sensors bounce a light beam and, based on how the beam returns to them (by bouncing off objects), the robot will know more about its environment. Some will use ultrasonic sensors (like bats) and these work in the same way – just using sound waves instead of light.
There are four basic types of sensors that you’ll find on a robot vacuum cleaner
- Cliff sensors are at the base of the machine and these determine the distance between the ground and the bottom of the robotic vacuum cleaner. The name tells us exactly why they are there. These sensors tell the bot when it reaches a sheer drop, like the edge of a step.
- Impact sensors work mechanically (not using a beam). These are commonly called bump sensors because they tell the bot when it has bumped into something.
- Wall sensors detect a continuous straight line, like a wall that the robotic vacuum cleaner has to follow.
- Optical encoders are the most sophisticated sensors found on a robotic vacuum cleaner and the cheaper models won’t always have this technology. These sensors track the movement of the wheels. By doing this, the bot knows how far it has traveled and how many turns it has made. This information allows the bot to always know where it is. This gives the robotic vacuum cleaner the ability to navigate.
Collecting information is one thing, but knowing what do with this information requires processing the information and making decisions based on the objective of the machine and the information it receives. The logic that scientists have used to create the artificial intelligence used in a robotic vacuum cleaner comes from a very unusual source – insects. Yes, that’s right, your robotic vacuum cleaner has the brain of a bug.
The thinking behind this is simple. If one takes a look at the behavior of insects like ants and bees, they fulfill a very simple purpose. Leave the nest, find food and bring it back. In order to do this, the insect needs a basic sensory input – its eyes and its feelers. This tells the bug where it’s going and where the food is. It needs to able to find the food and navigate back to the nest. That’s all the bug needs to figure out and it will keep repeating this task every day for the rest of its life.
In the same way, a robot vacuum cleaner has to leave the nest (its docking station), collect dirt and return to the docking station before the battery runs flat. By using the sensors listed above, the robotic vacuum cleaner navigates its way around the room without bumping into objects or falling down steps. It follows a logical path along walls and turning away from the wall to cover the room until it’s clean.
Of course, science never stops. So researchers have not stopped trying to improve the intelligence of the robotic vacuum cleaner. The more advanced (and more expensive) robot vacuum cleaners have intelligence that supersedes that of a bug. High-end robot vacuum cleaners have to main advancements – memory and the ability to interact with us.
The memory of a robotic vacuum cleaner records everything it does. So instead of always following a random path, avoiding obstacles, the intelligent bot forms a picture of the room based on its prior movement. These robot vacuum cleaners actually learn the layout of your home and are able to use this information to calculate a more efficient cleaning path. This enables them to clean your home faster and achieve much more on a single battery charge.
The ability of the robotic vacuum cleaner to interact with us comes in the form of an app that is installed on a smart device like a smartphone or tablet. Using blue tooth technology, the bot is able to send us information about itself and receive information from us. This technology is moving fast and phone apps can be updated at any time, so we will see many more interactive features being added as time goes by. At present, robot vacuum cleaners submit fairly basic information and this varies with different apps. Generally, they tell us how long they have cleaned for and we can check their recent history to see how well they are performing. They will also tell us if they get stuck, when the dust cup needs to be emptied, and other information that may affect their performance. We are able to control them remotely through the app and change their working schedules as needed.