Chapter 10: Adaptable robotics
The term "adaptable robotics" refers to a subfield of robotics that focuses on the development of robotic systems that are able to modify their hardware and software components in order to carry out a broad variety of tasks while simultaneously adapting to different situations. During the 1960s, robotics was first introduced into the industrial sector. After that, the discipline of adaptable robotics came into existence as a result of the requirement to create robots that possessed new types of actuation, adaptability, sensing and perception, and even the capacity to learn independently. The field of adaptable robotics has made remarkable strides forward thanks to significant developments such as the PUMA robot, research on manipulation, soft robotics, swarm robots, artificial intelligence, cobots, bio-inspired techniques, and other research that is currently being conducted. It is common practice to associate adaptable robots with their development kit, which is commonly employed in the process of developing autonomous mobile robots. In certain circumstances, an adaptable kit will continue to function normally even if certain components fail to function properly.
Modular design, machine learning, and sensor input are some of the strategies that are utilized by adaptive robotics systems in order to properly adapt to their surroundings. Because of this, they have transformed a number of different businesses and are able to address a wide range of real-world difficulties in the domains of medicine, industry, alien research, and experimentation. In the subject of adaptive robotics, there are still a great deal of obstacles to conquer, which presents potential for the sector to become more advanced.
A robot that is adaptable often possesses characteristics that set it apart from robots that are able to carry out their duty regardless of the environmental conditions that are present. Adaptability, sensing and perception, learning and intelligence, and actuation are the four concepts that adaptable robots use to differentiate themselves from other robots.
It is possible to define a robot as adaptive if it possesses qualities such as the capacity to learn, the ability to execute tasks that traditional robots are not capable of, and the ability to accomplish inherent safety and performance without compromise. The implementation of force control technologies, hierarchical intelligence, and various other novel ways are all viable options for achieving these capabilities. In 1994, John Adler came up with the idea for the cyberknife, which is a robotic surgery system that is capable of doing medical procedures with an extremely precise level of precision. This invention is an example of such adaptations.
The information about the environment that is gathered through peripherals is the subject of intelligent processing in adaptive systems. This data can be processed by AI systems, which can then alter task primitives in accordance with the data, resulting in customized action. The Canadarm 2 was flown to the International Space Station in 2001 and performed an important part in the maintenance of the station. It does this by utilizing data from peripherals to adapt the ISS to changes in the environment that occur within it.
Through the use of artificial intelligence, machine learning, and deep learning, systems are able to acquire knowledge about the environment in which they operate and grow increasingly intelligent as they carry out their tasks.[12] [12] Sojourner, a robot that was dispatched to Mars in 1997, was equipped with an internal computer that enabled it to adjust to unanticipated events and barriers even with minimum data. This provided a precursor to the incorporation of artificial intelligence into adaptable systems. The victory of IBM's Deep Blue computer against Garry Kasparov in a game of chess, which was a benchmark for the ability of robotic artificial intelligence to plan and respond, occurred later that year.
In robotic systems, the ability to move is made possible by the actuator. The majority of the time, adaptable actuators are designed to work in reaction to changes in their surrounding environment, such as variations in temperature, which can cause the actuator to change its shape. This results in a change in functionality. Especially in soft robotics, where external inputs can change the shape of an actuator, so creating mechanical energy, it is possible to achieve self-powering (untethered) actuation. Rodney Brooks has been responsible for the creation of Ghengis, a hexapedal robot that is capable of negotiating challenging terrain. In order to achieve mobility, the Hexapedal model makes use of six actuators. This model has maintained its prominence with contemporary hexapedal versions such as the Rhex.
Each of the kits includes an open-source software platform that has been customized to perform a variety of standard robotic operations. The kits also include typical robotics gear that can readily connect with the program (infrared sensors, motors, microphone, and video camera), which expands the capabilities of the robot. These components are packaged together with the software.
It is possible that the process of altering a robot to accomplish a variety of capacities, such as collaboration, might consist of nothing more than the selection of a module, the interchange of modules, teaching the robot through software, and carrying out the instructions.
The Venus flytrap serves as the foundation for the developing field of robotics using soft grippers, which is a subfield of the evolving field of adaptive robotics. Enveloping and pinching grasp modules are provided by two soft robotic surfaces at the same time. The grasping capacity of this technology is evaluated in a number of different contexts in order to assess the effects of a wide range of objects, faults in object position, and the installation of soft robotic surfaces. It is possible to exercise untethered actuation, particularly in soft robots that are constructed with liquid crystal polymers, which belong to the family of stimuli-responsive materials that have a two-way shape memory effect. Because of this, liquid crystal polymers have the potential to generate mechanical energy by transforming their shape in response to stimuli from the outside world, which is referred to as untethered actuation.
Robots that are built to operate in the outdoors and are able to adjust to different environments and challenges. It is possible for modular robots to transform into a variety of shapes in order to cross terrain since they are made in the form of a chain of independent modules that are joined together by simple hinge joints. Spider, serpentine, and loop configurations are examples of some of the formations that can be found here.
A subfield of robotics that applies the concept of swarm intelligence to groups of simple robots that are all the same. For the purpose of determining their movements in reaction to external cues, swarm robots are programmed to follow algorithms, which are typically meant to resemble the behavior of real animals.
Through the utilization of live tissues or cells, biohybrid robotics is able to give machines with capabilities that would be difficult to do in any other way. For instance, certain biohybrid robots have been able to move their bodies thanks to the utilization of muscle cells. There are several instances in which swarm robotics and biohybrid technology are combined, particularly in the field of medicine.
The qualities that adaptable robotics possess have made them suitable to a wide range of sectors, including but not limited to the fields of medicine, industry, and experimentation.
When it comes to translating human motion skills to robots, one approach that can be utilized is learning from demonstration. The major objective is to find significant movement primitives, which are considered to be significant movements that humans do, through demonstration and then recreate these motions in order to adapt the robot to that motion. There have been a few instances in which robots have been unable to adapt abilities that they have learnt through demonstration to new surroundings. This is a change from the scenario in which the robot was given initial demos. These problems with learning from demonstration have been solved by a learning model that is based on a nonlinear dynamic system. This model encodes trajectories as dynamic motion primitives, which are comparable to movement primitives but are significant movements that are represented by a mathematical equation. The variables in the equation change as the environment changes, which results in a change in the motion that is performed. It has been demonstrated that the trajectories that are recorded by these systems are applicable to a wide range of settings, which enables the robots to perform more effectively in their respective domains. The advancement of robots in areas where precision is of the utmost importance, such as surgical environments, has been made possible by the accumulation of knowledge through demonstration.
When it comes to the realm of medicine, SAR technology is centered on the collection of sensory data from wearable peripherals in order to determine the user's current state of being. Because of the information that is acquired, the machine is able to give tailored monitoring, incentive, and coaching for the purpose of helping with rehabilitation. To the point Physical human-robot interaction (HRI) and interfaces between humans and robots make it possible to perform activities such as recording the motions of a surgeon in order to infer their intent, determining the mechanical properties of human tissue, and other sensory data that can be utilized in applications related...
