Rachel Liddiard: Autonomous robot: Robot Vacuum cleaner

There are many different types of autonomous vacuum cleaner but they mainly fall into two categories: the ‘seeing’ robots and the robots which ‘touch’ obstacles by bumping into them to find their ways around rooms.

All the vacuums are compact so they can fit under most furniture and circular in shape which makes it easy to turn.

When the first generation of robot vacuum cleaners came on the market there were quite basic and some needed the size of the room to be programmed into them. They would frequently get ‘stuck’ around the house and needed a person to manually charge the battery up when it was flat.

Roomba

The first type of robot vacuum cleaner is the brand Roomba and it uses iRobot's AWARE(tm) Robotic Intelligence System which basically means that it can make decisions for its self and doesn’t need a lot of human input. It does this by using many different sensors to assess the environment around it which then sends the information to its microprocessor that then sends out an action for the robot to perform.

When the cleaning button is pressed on this robot the first thing it does is to check how big the room is using an inferred sensor to time how long the wave takes to bounce back. Once the robot has a rough estimate it starts to progress around the room.

It has two infrared sensors on the underside so the robot can avoid ‘cliffs’, for example stairs. The sensor senses when the infrared signal suddenly disappears this means that the robot is close to a ledge and so it reverses to avoid falling.

When the robot hits an object the bumper retracts which activates a mechanical sensor, the processor then chooses a random direction to go in to avoid the obstacle.

This robot can also find its charging docking station by ‘looking’ for the infrared signal given off by the docking station which it then follows and docks itself into the charger.

The main problem with this type of robot is that its pattern seems quite random and it is possible it doesn’t clean all of the room. It does not keep a record of where it has been and where it still needs to go.

Samsung

The second type of robot vacuum is the Samsung Hauzen. This used a camera to ‘see’ where it is going and therefore does not bump into walls or obstacles but ‘sees’ where they are. It is also a lot more methodical in its cleaning and goes back and forth in a zig zag motion across the room. The Samsung also has a ‘cliff’ sensor and can find and dock its self in its charger. This vacuum is a lot more expensive than the Roomba as the sensor and processes it uses are much more complex.

Conclusion

Even tho both these robots can make many desisions for themselves they still rely on the need for human input . They need to be told either by pressing a button or setting a timer, when to clean. And neither machine can fully understand where it has been and the Roomba sometimes gets stuck and need a person to help it out of trouble. These robots are not completely autonomous and no robot truly is, all robots need human input to either calibrate, start or switch off the machines.