Please enable JavaScript.
Coggle requires JavaScript to display documents.
3.9 - Coggle Diagram
3.9
Different types of robots
There are many types of robot platform, including arms, drones, vehicles, humanoids and various bioinspired designs such as crawling robots.
Industrial robots
Articulated robots (e.g. arms & snakes) with independent rotary joints
SCARA (Selective Compliance Articulated Robot Arm), rotary joints with parallel axis
Parallel (Stewart platform), linear actuators linked to platforms
Cartesian, linear actuators placed orthogonally for independent xyz motion
Service Robots
Humaniod robots
Wheeled robots and autonomous guided vehicles (AVGs)
Drones and unmanned aerial vehicles (UVAs)
Exoskeletons
Exoskeletons are an external structure with joints and links that go around the user’s body to aid in movement.
They are wearable machines, powered by electric motors, pneumatics, levers, hydraulics or a combination of technologies
Gopura and Kiguchi (2009) classified exoskeletons into the four categories below:
By Applied Segment
e.g. part of the human body where the exoskeleton is attached
By Actuators
Electric motors, pneumatic, hydraulic actuator, mechanical.
By Function
e.g. rehabilitation, motion assist, human power, haptic interaction
By Power Transmission
Gear-driven, cable-driven, linkage mechanism, combined
Cloud robotics
Artificial intelligence is primarily based on machine learning. It builds on algorithms and neural networks that learn using information from large amounts of data.
Data centres distributed around the world house our data. A typical data centre costs more than $1Bn, is several hundred thousand square feet in size and houses circa 100,000 servers in racks.
5/6G allows high bandwidth connection to data centres with low latency, allowing IoT devices to connect seamlessly to one another and the cloud.
Moore’s law is coming towards its end and there a limit to how much computing power can be put on a mobile robot because of battery life.
Even if quantum computers deliver on their potential, the requirement for cooling down to millikelvins means that they won't be mobile anytime soon.
Introduction to Robotics
A robot needs to have four essential characteristics:
Sensing
: A robot must be able to sense its environment. Sensors are used to provide a robot with an awareness of their surroundings. These sensors are specific to light, touch, pressure, chemical, sound (including sonar), and radio waves (e.g. radar).
Movement
: A robot is required to perform the desired movements efficiently within its environment.
Energy
: A sufficient power source such as mains, battery, solar or wireless is required so a robot can power itself.
Intelligence
: A robot must be programmed to perform tasks that it is required to do.
Controlling robots
Robots can be controlled in several different ways:
The movements of a robot are pre-programmed.
The robot plays back a sequence of poses and tasks
An example of this would be a six-axis arm in a factory, programmed to repeat the same steps over-and-over again.
Robot arms could be programmed in
ABB's RobotStudio
software
RobotStudio™ can be used to design and programme a complete manufacturing cell offline, simulating the coordinated motion of multiple robots prior to downloading the commands to the robot controllers on the factory floor.
Types of control:
Wireless Control
Allows a robot to be controlled wirelessly. Telepresence has made this more flexible than ever before, since a human user can see in 3D remotely and operate a robot with wearable sensors.
An example of this is motion capture gloves with haptic force-feedback.
Cloud Control
This allows the robot to operate remotely through cloud computing technologies.
Cloud robotics enables access to vast databases of information (e.g. photographs for object recognition, maps for motion planning) as well as computing power.
Standalone Control
his allows a robot to function without the assistance of a computer. A computer is used initially to set up and program the robot, but afterwards, the computer is no longer required
Artificial intelligence (AI) and robotics
Enhanced capabilities include:
Grasping
Using AI, robots can grasp items which they have never seen before. They can also determine the best position and orientation to grasp an object.
Data
AI helps robots to understand physical and logistical data patterns. This means a robot can respond proactively and act accordingly.
Computer Vision
AI can help robots to detect and recognise objects – potentially ones that they have never seen before.