How robots work The working principle and process of robots

Autonomous robots can act on their own, without relying on any human controller. The basic principle is to program the robot to react in a certain way to external stimuli. A very simple collision reaction robot is a good example of this principle.

This robot has a collision sensor that detects obstacles. When you start the robot, it zigzags in a roughly straight line. When it hits an obstacle, the impact acts on its collision sensor. Each time a collision occurs, the robot is programmed to back up, turn right, and then continue forward. In this way, the robot changes its direction whenever it encounters an obstacle.

Advanced robots will use this principle in more sophisticated ways. Roboticists will develop new programs and sensor systems to create robots that are more intelligent and more perceptive. Today's robots can perform well in a variety of environments.

Simpler mobile robots use infrared or ultrasonic sensors to sense obstacles. These sensors work similarly to the echolocation system of animals: the robot sends out a sound signal (or a beam of infrared light) and listens for the reflection of the signal. The robot calculates the distance between it and the obstacle based on the time it takes for the signal to reflect.

More advanced robots use stereo vision to see the world around them. Two cameras give the robot depth perception, while image recognition software gives the robot the ability to determine the location of objects and identify various objects. The robot can also use microphones and smell sensors to analyze the surrounding environment.

Some autonomous robots can only work in limited environments that they are familiar with. For example, a lawn-mowing robot relies on buried landmarks to determine the boundaries of the lawn. And a robot designed to clean an office needs a map of the building to move between different locations.

More advanced robots can analyze and adapt to unfamiliar environments, even to areas with rough terrain. These robots can associate specific terrain patterns with specific actions. For example, a rover robot will use its vision sensors to generate a map of the ground ahead. If the map shows a rough terrain pattern, the robot will know that it should take a different path. This system is very useful for exploratory robots working on other planets.

One candidate robot design is loosely structured, with an element of randomization. When it gets stuck, it moves its appendages in all directions until its actions work. It does this by tightly coordinating force sensors and actuators, rather than having a computer program directing everything. It's similar to an ant trying to get around an obstacle: the ant doesn't seem to make a quick decision when it needs to get around an obstacle, but instead keeps trying different approaches until it gets around the obstacle.

3. Homemade Robots

In the final sections of this article, we'll look at the most notable areas of the robotics world: artificial and research robots . Experts in these fields have made great advances in robotics over the years, but they are not the only ones making robots. For decades, a small but passionate group of hobbyists have been building robots in garages and basements around the world.

Homemade robotics is a burgeoning subculture with a sizeable presence on the Internet, where amateur robotics enthusiasts assemble their own creations from a variety of commercial robotics kits, mail-order parts, toys and even old VCRs.

As with professional robots, there are many types of homemade robots. Some robotics enthusiasts who work on weekends build very sophisticated walking machines, while others design household robots for themselves. Still others are keen on building competitive robots. The most familiar competitive robots are remote-controlled robot warriors, like those you see on the show "BattleBots." These machines are not "real robots" because they do not have reprogrammable computer brains. They are just souped-up remote-controlled cars.

More advanced competitive robots are controlled by computers. For example, soccer robots play small soccer games with no human input at all. A standard robot soccer team consists of several individual robots that communicate with a central computer. This computer "sees" the entire field through a camera and distinguishes the ball, the goal, and its own and opposing players by color. The computer processes this information all the time and decides how to direct its team.

Adaptability and versatility

The personal computer revolution was marked by its remarkable adaptability. Standardized computers and programming languages ​​made it possible for amateur programmers to build computers for specific purposes. Computer parts were a bit like craft supplies, and their uses were countless.

Most robots to date are more like kitchen appliances. Roboticists build them to do one specific job. But they don't adapt very well to completely different applications.

That’s changing. A company called Evoluon Robotics, which is pioneering adaptable robotics hardware and software, is hoping to carve out a niche with an easy-to-use “robotics developer kit.”

The toolkit has an open software platform that is dedicated to providing a variety of commonly used robotics functions. For example, roboticists can easily give their creations the ability to track objects, listen to voice commands, and navigate around obstacles. From a technical perspective, these features are not revolutionary, but what is unusual is that they are integrated into a simple software package.

The kit also comes with some common robotics hardware that can be easily integrated with the software. The standard kit provides some infrared sensors, motors, a microphone and a camera. Roboticists can assemble all these parts using a reinforced mounting kit that includes some aluminum body parts and durable wheels.

Of course, this kit isn't for making mediocre creations. At more than $700, it's by no means a cheap toy. But it's a big step toward a new kind of robotics. In the not-too-distant future, if you want to build a new robot that can clean your house or take care of your pet while you're away, you might be able to do it by writing a B program, which could save you a ton of money.

4. Artificial Intelligence

Artificial intelligence (AI) is undoubtedly the most exciting area in robotics, and undoubtedly the most controversial: everyone agrees that a robot can work on an assembly line, but there is disagreement as to whether it can be intelligent.

Just like the term "robotics" itself, "artificial intelligence" is difficult to define. The ultimate artificial intelligence is the reproduction of human thought processes, that is, an artificial machine with human intelligence. Artificial intelligence includes the ability to learn any knowledge, reasoning, language, and the ability to form one's own opinions. Robotics experts are still far from achieving this level of artificial intelligence, but they have made great progress in limited areas of artificial intelligence. Today, machines with artificial intelligence can already imitate certain specific elements of intelligence.

Computers already have the ability to solve problems in limited areas. The execution of problem solving with AI is complex, but the basic principle is very simple. First, an AI robot or computer collects facts about a situation through sensors (or human input). The computer compares this information with stored information to determine what it means. The computer calculates various possible actions based on the information collected and then predicts which action will be most effective. Of course, a computer can only solve the problems it is programmed to solve, and it does not have analytical ability in the general sense. A chess-playing computer is an example of such a machine.

Some modern robots also have limited learning capabilities. A learning robot is able to recognize whether a certain action (such as moving a leg in a certain way) achieved a desired result (such as getting around an obstacle). The robot stores this information, and the next time it encounters the same situation, it tries to make the movements that will successfully handle it. Again, modern computers can only do this in very limited situations. They cannot collect all types of information like humans can. Some robots can learn by imitating human movements. In Japan, roboticists taught a robot to dance by showing it dance moves.

Some robots have the ability to communicate with people. Kismet, a robot made by MIT's Artificial Intelligence Lab, can recognize human body language and the tone of voice and respond accordingly. The authors of Kismet were interested in the way adults and infants interact with each other, which can be completed with only tone of voice and visual information. This low-level interaction can serve as the basis for human-like learning systems.

Kismet and other robots built by MIT's Artificial Intelligence Laboratory use an unconventional control structure. Instead of having a central computer controlling all actions, these robots have low-level actions controlled by low-level computers. Rodney Brooks, the project director, believes this is a more accurate model of human intelligence. Most of the actions of humans are automatic, rather than determined by the highest level of consciousness.

The real challenge of artificial intelligence is to understand natural intelligence. Unlike building an artificial heart, scientists do not have a simple and concrete model to refer to when developing artificial intelligence. We know that the brain contains tens of billions of neurons, and we think and learn by establishing electronic connections between different neurons. But we don't know how these connections achieve high-level reasoning capabilities, or even the implementation principles of low-level operations. The brain seems too complex to be understood.