Artificial intelligence (AI) is a very important aspect of robotics and perhaps, very controversial. A lot of people such as Elon Musk, Mark Zuckerberg and Jeff Bezos are wondering if in fact a machine can think like human what the implications are of that happening. The ultimate AI will be able to recreate the human thought process such as the ability to think, learn and create. Artificial intelligence in robotics is still very limited despite the tremendous progress made in recent years and at this stage have a long ways to go before a robot can be humanlike in its “thinking”.
What is a Robot?
The term “robot” was first used in the 1921 Karel Capek’s play “R.U.R. Rossum’s universal robots.” The word robot is actually Czech in origin and is based on the Slavic word “robota” which roughly means being in servitude and doing and doing repetitive boring tasks.
Generally speaking, robots are mechanical devices that manipulate the physical world by performing simple tasks. Robots are equipped with sensors to perceive their environment and effectors to physically assert force on the environment. We can think of robots as the hardware like a computer and the AI is the software like the operating system.
A high level overview of AI and Robotics
Telemanipulators and numerical control are the two enabling technologies that robotics are built on. At some level these two have to be controlled or managed by something. That might be simple software in the robot or it might be a person.
Telemanipulators are machines that are remotely controlled which usually consists of a gripper and an arm. The human uses a control device to regulate the movements of the arm and gripper. The first telemanipulators have been used to regulate and control radio-active materials.
Numeric Control makes use of coordinate system to precisely control machines. The first programming language for machines (APT: Automatic Programmed Tools) was developed when Numeric control was first used in 1952 at the MIT.
The first programmable telemanipulator was developed from the combination of these two enabling technologies. The robots installed in industrial facilities like car construction plants were built using the principle of programmable telemanipulator.
Mobile robots came into the picture when there was need for the automation of transportation in production processes and autonomous transport systems. The desire to automate transportation in production processes led to driverless transport systems used on factory floors to navigate objects to different stages in the production stages in the late seventies. In recent years, new forms of mobile robots (such as the insectoid robots with many legs) have been built to model nature.
Sensors are the tools robots use to relate to their environment. The sensors are of two types: active and passive sensors. Passive sensors (for example: camera) capture signals that are produced by other sources in the environment. Active sensors (for example: radar laser and sonar) emit energy into the environment that are used to gather vital information about the environment.
These are what the parts of the robots that are used in manipulating the environment, move and change their body shapes.
For mobile robots, the mechanisms for locomotion, including tracks, wheels and legs, are a special group of effectors.
4. Power Sources
The effectors in robots need a power source to be driven. For both manipulator actuation and locomotion, the most popular mechanism is the electric motor. Other possible power sources include: pneumatic actuation which uses compressed gas and hydraulic actuation which uses pressurized fluids.
5. Data Connection
Most robots have wireless networks for digital communication. These modules are used for communication between robots or as a backlink to the home station of the robots
Types of Robots
1. Manufacturing and Industrial Robots
Like assembly line work in the automobile industry, hardworking robots are traditionally being used to replace human workers in areas of difficult labor. Other examples of hard working robots include:
- Harvesting machines
- A robot that removes paint from large ships
- Mining transportation
In spite of the fact that most autonomous robots still need environmental modifications to navigate their way, they are already used widely. Examples of the current transporters available are:
- Autonomous helicopters used in delivering goods to remote areas
- Medication and food transport systems in hospitals
- Container transporters used in loading and unloading cargo ships
3. Hazardous Environments
These are the robots used in dangerous, hazardous or nuclear environments where humans can’t easily go. Some examples include:
- At the World Trade Center after 9/11
- Exploring the Fukushima Nuclear disaster area created by the tsunami
- Robots have helped in cleaning up nuclear waste in Chernobyl
- Robots are constantly being used in cleaning ammunition and mines across the globe.
- Including sending robots to Mars and deep into oceans for exploration
4. Domestic and Professional
These are new models of robots. They are built to help in everyday life which includes anything from sweeping the floor to assisting with surgery. This even includes the robot kits for learning and education that kids can take advantage of.
The current state of Technology
The complexity of these robots has to be managed. That is inevitably done through a combination of environmental modifications (e.g. a specific stand for a robot arm), some level of software and of course humans.
Examples of AI and robotic technology being combined today include:
- Self driving cars – The robot is the car and the AI drives it…
- Auto Pilot for Airplanes – Yes that’s an AI flying the playing – it was one of the earliest uses of Artificial Intelligence
- Robot Vacuum Cleaners – These vacuum cleaners have a small AI driven “brain” that allows them to successfully complete their tasks. For example some of Sharp Electronics industrial vacuum robots use their COCORO AI Engine.
The goal of combining AI with robot technology is to allow the AI to integrate with the physical robot to manage and control its function. This is not unlike our own brains controlling our bodies. But its also more than control. The goal is to allow the entire robotic mechanism to make decisions. Perhaps in a factory environment that’s deciding what task to do next for optimal efficiency versus just taking the next task. Or for the medical industry a service robot can make a decision on when to call for help.