Robotics and Automation in Industry 4.0

Nanorobots: Types, Principles and Applications

Parkarsh Kumar¹, Santosh Kumar Das¹, *

¹ Sarala Birla University, Ranchi, India

Abstract

Nanobots or Nanorobots are one of the emerging applications in both nanotechnology and robotics. These bots are programmed to carry out specific applications for a specific purpose. Owing to their properties such as smaller volume, efficiency and accuracy, nanobots are being explored in different fields of study, especially nanomedicine, automation, drug delivery, chemistry, aerospace and others. These bots can be programmed and explored in such a way that they can be used to repair the specific target in the body, which is impossible using bare hands. In this chapter, we are going to explore such types of applications and their principles.

Keywords: Application of nanobots, Nanobots, Types of nanobots.

* Corresponding author Santosh Kumar Das: Sarala Birla University, Ranchi, India; E-mail: [email protected]

INTRODUCTION

In 1959, Professor Richard Feynman, through his talk Plenty of Room at the Bottom, laid the fundamentals of nanotechnology [1]. In one of the chapters, he discussed molecular machines, the application of which includes nanorobotic surgery, drug delivery, and others. As suggested by Professor Feynman, the miniaturization of robotics has been explored, and advancements have been made in this field over time. These are types of nanodevices whose application can be explored in many fields, but currently, it is mostly being explored in the field of biomedicals. One of the first studies on nanobots was conducted by Robert Freitas, which is based on the principle of molecular manufacturing [2]. This nanodevice was created to function like an artificial erythrocyte.

Compared to natural red cells, these artificial devices were capable of transporting a similar amount of carbon dioxide and 236 times more oxygen to tissues. These devices were fabricated using hard diamondoid materials and consist of different types of chemicals, pressure and thermal sensors to monitor complex responses and behavior. Since then, nanobots have evolved in multidimensional fields with increasing applications. Mavroidis et al. [3] explained theevolution of fully

functional autonomous nanorobots and the integration of different components into a system. Owing to its small size, durability, self-replication and mobility, multiple applications of nanobots have been developed to solve problems related to the environment, agriculture, biology and medicine. The design of different nanorobots is inspired by different kinds of species present in the environment to perform specific operations. These nanobots seem to be very promising in areas of invasive surgery, drug delivery, contamination screening, controlling pests and plant viruses, antimicrobial activity, breaking blood clots, parasite removals, and others [4-6]. In order to scale up, researchers across different organizations are trying to overcome limitations in terms of high-quality design, cost analysis, biocompatibility, interfacing with blood, which is viscous, and control issues. Overcoming the challenges will help in achieving unlimited applications especially in the field of biomedical. The current trend of nanorobots in precision medicine, which includes surgeries, therapy, imaging and diagnosis, is being studied [7]. These bots will have the ability to locate specific positions in the body and deliver the precise dosage. With noninvasive medication and various groundbreaking innovations, it is expected that the market of nanobots will exponentially increase to 21.45 billion dollars in 2030 from its present value of 6.19 billion dollars.

Fig. (1))

Types of nanorobots.

TYPES

Further, we are going to discuss different types of nanobots based on a few parameters such as design, nanorobotic devices, actuation types and manufacturing approaches. These are just a few parameters, which can be further subdivided into many categories. Some types of robots based on the subdivision of parameters can be similar in nature. Nanobots classification and their respective subcategories are shown in Fig. (1).

On the basis of design: Based on the applications, many designs of nanobots have been proposed, but they are generally classified into three categories on the basis of design. Arvidsson et al. described them as Helices, DNA Nanobots and Nanorods [8-11]. They are discussed in detail in further sections.

On the basis of actuation: In general, the movement of nanobots takes place in a medium having a low Reynolds number, where inertial force is negligible as compared to viscous force. This makes it difficult to operate, and hence nanobots need to be uninterruptedly powered for actuation. Since it is very difficult to load energy sources on nanorobots, researchers have started focusing on actuation technologies for nanorobotics. Xu et al. [12] summarized different types of actuation techniques into two sub-categories, which are self-actuation and external field actuation. In order to actuate nanobots using a variant magnetic field, as explained by Dreyfus et al. [13], it is subject to get a constant and oscillating magnetic field along the horizontal and perpendicular direction, respectively. Later, many other techniques, such as U-shaped, fish-like structure, freestyle swimmer, sea star and other types of magnetically actuated nanobots, have been proposed for different applications. Although there is a safety issue related to high-intensity magnetic fields, strong pungent power and no harm to the biological system make it viable to experiment on. We know that an electric and magnetic field can be transformed into each other, so robotic systems with certain conditions and without complicated mechanical parts were proposed, which could utilize an external electric field for actuation. For actuation, some researchers use the finite element technique to produce a magnetic field using coil assembly. The concept of quad nanopore devices is also used for the purpose of actuation. Other types of external field actuation include light field actuation and acoustic actuation. Experiments were conducted to establish a relationship between laser intensity and the speed of nanoswimmers for biomedical and soft matter applications. Based on such principles, needle-shaped nanoswimmers, electrochemical actuators, and photoresponsive DNA signaling molecules have been designed over time. On the other hand, researchers used the principle of ultrasonic bubble-gathering effect for the movement of robots. Rod-shaped nanobots that can rotate, arrange, suspend and assemble in the ultrasonic range are fabricated. Some researchers achieve cell manipulation and actuation using acoustic radiation force in standing waves. In the case of self-actuation methods, the static balance of nanorobots gets disturbed for the purpose of movement. Before breaking the static balance, an asymmetrical chemical field needs to be created. Whitesides et al. [13] was one of the first researchers who performed self-actuation in robots by converting chemical energy into kinetic energy using the principle of oxidation-reduction. Later, tubular and platinum-based V-shaped robots were developed for self-actuation. Some researchers even worked on a hybrid Au/Pt (z-shaped) self-actuation system for targeted drug delivery. These chemical self-actuation devices were biocompatible, but these methods also have safety issues, short action time and nontoxic urea. Biological self-actuation can be used to overcome the issue of safety. These types of nanobots have been developed by taking inspiration from nature. Flagella and Magnetotactic bacteria give promising results for actuating force and smart nanobots. Droplets of different bacteria were placed in a solution of glucose and water to generate movement with the help of ethylenediaminetetraacetic acid [14].

On the basis of devices: Nanobots can be broadly classified into four categories. Mavroidis et al. [3] mentioned these four systems as Nanomanipulators, BioNano Robotics, Magnetically Guided Nanorobotics and Bacteria-based Nanorobotics. The microscopic viewing system that helps in viewing small-scale objects is termed the nanomanipulator. Scanning probe microscopes are used to enlarge images, which helps in a better understanding of devices and manipulation at atomic scales. Some biological elements, such as DNA and proteins, are capable of generating force or signal at the cellular level. Devices made from these elements are called nanocomponents, and the robotic system is called the bio nanorobotic system. Developed in 2003, these systems are now capable of transformation into motors, sensors, joints and other components of robotic systems. It has been hard to develop full-scale nanorobots, but with the development of the ferromagnetic material-based magnetically guided nanorobotic system, nanobots with six degrees of freedom (DOF) can be realized and external magnetic force can be used for propulsion and actuation. Another type, called bacteria-based nanorobotic system, is developed by taking information from nature [15]. Although it has similarities with both magnetic and bio nanorobotic systems, it differs in terms of design, control and guidance.

On the basis of manufacturing: Based on the approaches for manufacturing nanorobots, they can either be classified as NUbots or bacteria-based [16]. The bacteria-based approach uses flagellum and electromagnetic field for propulsion and control of motion by taking inspiration from microorganism bacterium such as E. coli. It is the same as what we have discussed in other classifications of nanobots. Nubots are also called nucleic acid robots. They are a type of molecular machine thatis fabricated and activated using molecules, proteins and DNA.

These are the limited classifications. Further subcategories of nanobots can be created by conducting a comparative study on nanobots on the basis of different parameters related to them.

Applications and Recent Developments

We have discussed different types of nanobots, their principle and their fabrication approaches. It is yet to be fully scaled, but based on the approaches, it has been slightly used in different parts of devices, especially in biomedical devices. Experts from different fields of study came together to make advancements in nanobots, which is certainly the future of precision medicine, non-invasive diagnostics, environmental science, agriculture, military, space and others. Most of the devices are still in the preliminary phase of development, and lots of research and trials need to be performed in order to make them commercially and economically viable.

With research advancement and commercialization, full-scale, economically viable nanobots can be developed to solve problems related to medicine, like cancer treatment, by replacing damaged neurons or by transferring specific drugs to specific sites, thereby safeguarding patients from the harmful process of chemotherapy [4] [13]. It can also be used in the body to monitor abnormalities and provide noninvasive ways to treat them. Using this kind of surveillance technique, it can help in checking the contents of blood and provide warning about any kind of possible diseases. Other medical problems that can be monitored and diagnosed are blood clots, body clots, diabetes, gout, and others.

With rapid development in the field of materials and biochemistry, the future of nanobots is near, but challenges remain in terms of removal of toxicity, proper miniaturization of devices, compatibility, energy harvesting, positional nano assembly, noise interference and dynamics of robotic systems [17]. Some decision making, AI-powered tools and machine learning can also be integrated in the future for better outcomes [18-23]. Some challenges are also present during the designing of nanorobots, such as high complexity, interface design, cost, etc. There are several post-quantum cryptography techniques available to implement security in the nanorobots [24-27]. They help to reduce several attacks in the context of different applications. Lightweight cryptography also helps to protect against some attacks [28-34].

CONCLUSION

The need to describe rarefied flows in a variety of applications that involve micro and nanoscale pores or channels has stimulated the development of a multitude of theoretical and numerical methodologies. A comparison of the predictions of the mesoscopic approaches with those of the Navier-Stokes formulation complemented by a non-slip flow condition revealed that the latter can be safely employed for very small Knudsen number values only. The DSMC method was used to determine the flow field in different types of porous media and processes.

REFERENCES

Robot Path Planning in a Dynamic Environment Using Deep Q-Learning

Rifaqat Ali¹, *, Preeti Chandrakar²

¹ Department of Mathematics and Scientific Computing, National Institute of Technology Hamirpur, H.P., India

² Department of CSE, National Institute of Technology Raipur, India

Abstract

Robot path planning is a necessary requirement for today’s autonomous industry as robots are becoming a crucial part of the industry. Planning a path in a dynamic environment that changes over time is a difficult challenge for mobile robots. The robot needs to continuously avoid all the obstacles in its path and plan a suitable trajectory from the given source point to the target point within a dynamically changing environment. In this study, we will use Deep Q-Learning (Q-Learning using neural networks) to avoid the obstacles in the environment, which are being dynamically created by the user. The main aim of the robot is to plan a path without any collision with any of the obstacles. The environment is simulated in the form of a grid that initially contains information on the starting and the target location of the robot. Robots need to plan an obstacle-free path for the given points. The user introduces obstacles whenever he/she wishes during the simulation to make the environment dynamic. The accuracy of the path is judged by the path planned by the robot. Various architectures of neural networks are compared in the study that follows. Simulation results are analyzed for the evaluation of an optimized path, and the robot is able to plan a path in the dynamic environment.

Keywords: Artificial neural network, Deep learning, Neural networks, Reinforcement learning, Robot.

* Corresponding author author Rifaqat Ali: Department of Mathematics and Scientific Computing, National Institute of Technology Hamirpur , H.P., India; E-mail: [email protected]

INTRODUCTION

Nowadays, with the advancement of technology and a tremendous boom in the field of robotics, robots are taking over most mundane tasks that do not require much cognitive ability and can be performed easily by unskilled laborers. In 2010, Chen et al. [1] showed that robots used to serve and cater to customers in a human-interactive environment in restaurants. Later, Chen et al. [2] also proposed

the vacuum robots used for cleaning. In 2015, multiple robots communicated with each other over a shared network and carried loads from one fixed location to another fixed location, as shown by Das et al. [3]. Autonomous mobile robots have found their way into industries, military research programs, and many scientific research programs. The Schemes [4-6] present additional information on robots used in human-friendly environments like education, entertainment, surveillance and many more. As seen, robots continuously keep evading new human-centric and non-human-centric environments to find vast applications. Robots need to be as autonomous as possible i.e., they should possess the ability to navigate, find paths and perform various activities without the intervention of any kind to accomplish all such tasks. All the activities that are performed by autonomous mobile robots require the design and implementation of a path planning and path optimization algorithm at its core.

Robot path planning, being at the crux, plays a vital role in the design, implementation and manufacturing of autonomous robots. If implemented effectively, path planning significantly improves the following parameters in an autonomous robot:

Robot Accuracy: The algorithm must produce next to zero errors because the robot is expected to work in human-friendly environments where the slightest errors can be catastrophic. Hence, there is a need to test this out and make sure the robot is accurate enough that it stands firm on its capabilities.

Task Repeatability: Once the implementation of the path planning algorithm is complete, including the training phase of the robot, it can perform the same task hundreds of times with high accuracy and effectiveness without any slack.

Product Quality: Based on accuracy and repeatability, we can directly conclude that our robot has been trained well and can assume a result that will depict the high-quality work of our robot.

One can define path planning as the continuous locomotion of a robot from the initial/starting configuration to the final/goal configuration in an unknown environment. The environment in which the robot moves consists of free space and obstacles. The robot needs to ensure that it does not collide with any obstacle during its course. Path optimization deals with finding the path with the least cost and distance and which consumes the least amount of time. Fig. (1a) shows an example environment with obstacles and defined configurations for start and goal states. Fig. (1b) shows an example of a path planned in the shown environment.

Structure of the Paper

This article is organized into 7 sections. Following the introduction, Section 2 presents different approaches and work, which were used by various researchers in the past to solve the problem. Also, the section ends with a description of the approach used in this article. Section 3 of the article describes the methodology used to solve the problem in addition to the nuances of the algorithm. Section 4 discusses the experimental setup and the various technologies used, along with the various parameters of the algorithm used during the experiment. Section 5 analyses the results and the performance of the experiment with the help of some graphs. The article comes to an end by presenting a conclusion for the experiment in Section 6.

Fig. (1))

(a) Environment with obstacles and start and goal configurations, and (b) one out of many possible paths from the start to the goal configuration.

RELATED WORKS

In recent decades, many approaches have been devised by researchers to find a solution to the problem of path planning and optimization for autonomous mobile robots. These approaches range from the use of graph-based algorithms to the use of contemporary neural networks to implement deep learning and reinforcement learning algorithms. Some of these approaches and algorithms have been briefly described as follows:

Graph-Based Approach

In this type of approach, the environment is considered as a connected network consisting of nodes and edges similar to a graph data structure. The nodes that constitute the obstacles are considered forbidden since the robot is not allowed to use these nodes while ascertaining a path from the initial to the goal state. All other nodes form the free space of the environment and are traversable as the robot moves. Classical graph traversing algorithms, such as the Breadth-First Search (BFS) and the Depth First Search (DFS), are fruitful in determining solutions to the path planning problem.