19-10-2010, 02:41 PM
[attachment=6573]
[attachment=6574]
Nanosystem Design with Dynamic Collision Detection for Autonomous Nanorobot Motion Control using Neural Networks
INTRODUCTION
The starting point of nanotechnology to achieve the main goal of building nanoscale systems is the development of autonomous molecular machine systems. The presented paper describes the design and simulation of autonomous multi-robot teams operating at atomic scales with distinct assembly tasks. Teams must cooperate with each other in order to achieve a productive result in assembling biomolecules into larger biomolecules. These biomolecules will be delivered to “organs†(into a set of predefined organ inlets), and such deliveries must also be coordinated in time.
Building patterns and manipulating atoms with the use of Scanning Probe Microscopes (SPM) as in Atomic Force Microscopy and Scanning Tunneling Microscopy [19] is a promising approach for the construction of nanoelectromechanical systems (NEMS) with 3D precision at up to 0.01 nm resolution. However, these manual manipulations require much time and at present such repetitive tasks give imprecise results when performed manually on a large number of molecules. Approaches for nano-planning systems have been presented [19] as a first step towards automating 2D assembly tasks in nanorobotics, and the possible use of artificial intelligence as the appropriate means to enable some aspects of intelligent behaviour for the control of nanorobots in molecular manufacturing automation has been discussed in the nano community [08]. Theoretical work in molecular manufacturing has emphasized the need for very small and very accurate manipulators which simultaneously have a wide range of motion to enable the task of assembling molecular components [10]. More recent work in the possible automation of nanoscale manipulation has produced a fully autonomous motion manipulator system capable of performing 200,000 accurate measurements per second at the atomic scale [20].
•Â 2. NANOMEDICINE
The principal focus in medicine is going to shift from medical science to medical engineering, where the design of medically-active microscopic machines will be the consequent result of the techniques provided from human molecular structure knowledge derived during the 20th (and the beginning of the 21st) century [11]. For the feasibility of such achievements in nanomedicine [11] two primary capabilities are required: fabrication of parts and assembly of parts. Through the use of different approaches such as biotechnology, supramolecular chemistry, and scanning probes, both capabilities had been demonstrated in limited fashion as early as 1998 [11]. Despite quantum effects which impose a relative uncertainty to electron positions, such objections are resolved by recognizing that the quantum probability function of electrons in atoms tends to drop off exponentially with distance outside the atom, giving atoms a moderately sharp "edge". Even in most liquids at their boiling points, each molecule is free to move only ~0.07 nm from its average position [11]. Recent developments in the field of biomolecular computing [01] have demonstrated positively the feasibility of processing logic tasks by bio-computers [14], which is a promising first step to enable future nanoprocessors with increasing complexity, and nanoscale information storage and data processing capacity, which could be considered as an indispensable component of a real autonomous nanosystem. Other advances in the sense of building biosensors [25] and nano-kinetic devices [24] have advanced recently too, which could be considered as well a prerequisite for making nano-automation feasible and enabling nanorobotics control and locomotion. Many classical objections to the feasibility of nanotechnology, such as quantum mechanics, thermal motions and friction, have already been considered and resolved [10]. The presented nanorobot will be required to perform a pre-established set of tasks in the human body as is a ribosome, which is a natural molecular machine system [11].
•Â 3. PROPOSED DESIGN
A multi-robot molecular machine system could be described as a system to perform molecular manufacturing at the atomic scale, whose constituent entities are capable of cooperating collectively. Three main design approaches for nano manipulation in the liquid or air environments are: robotic arm, Stewart platform and a five-strut crank model. For our experiments we chose nano-manipulation in a liquid environment, which is most relevant within the presented application in nanomedicine. It was demonstrated that computation is relatively cheap for macroscale robotic actuators while arm motion is relatively cheap for nanoscale robotic actuators. Thus the moment-by-moment computer control of arm trajectories is the appropriate paradigm for macroscale robots, but not for nanoscale robots [11]. For nanoscale robots, the appropriate manipulator control is often trajectory trial and error, also known as sensor based motion control [16].
Virtual Environment
Virtual Reality was used for the nanorobot design where the use of macro- and microrobotic concepts is considered a practical approach once the theoretical and practical assumptions here have focused on its domain of application. The design should be robust enough to operate in a complex environment with movement in six-degrees-of-freedom. Nanoscale object manipulation systems have been applied with the use of computer graphics for teleoperation. The requirements for such systems have been clearly established [23]. A starting point for our hypotheses and experiments was to consider the robot design derived from biological models and comprised of some basic nanoscale components such as molecular sorting rotors and a telescoping manipulator (robot arm) [10]. The robot design adopted concepts provided from underwater robotics [27] keeping in mind however the kinetics assumptions that the nanorobot lives in a world of viscosity, where friction, adhesion, and viscous forces are paramount and gravitational forces are of little or no importance [11]. The obstacles will be located in unknown positions (figure 1). The delivery positions that represent organ inlets requiring proteins to be injected are located in a well-known position for the nanorobot teams if these organ inlets are (or are not) scheduled for injection at time t, they will change their colours, indicating the opening or closing of the team A (blue nanorobots) and B’s (yellow nanorobots) delivery orifice, which will indicate for the agents if they could perform their delivery in the correct order (figure 2). The trajectories and positions of each molecule are generated randomly and each molecule also has a probabilistic motion acceleration. The nanorobot navigation uses plane surfaces (three fins total) and bi-directional propellers, which are comprised of two simultaneously counter-rotating screw drives for the propulsion [11]. Considering the liquid environment, a sonar approach seems to be the most appropriate choice of sensor device for nanorobots in nanomedicine [05], thereby for navigational purposes the blue cones shown in figure 3 represent regions that the robot’s sonar can “hearâ€. Scientific visualization techniques permit rapid and precise geometric analysis for a sonar classification system [04]. The nanorobot sensors report collisions and identify when an encountered object is an obstacle to be avoided or a molecule to be caught. While some molecules are being captured (figure 3), other molecules will be assembled internally by the robot arm.
3.2 Physically Based Simulation
The study of non-penetrating rigid bodies in virtual reality for dynamic constrained simulation is a field of research in computer graphics that has an enormous impact for physically based simulation and a large range of works in this field have achieved good results. Particularly in calculating motions of many objects that move under changing constraints and frequently make collisions, one of the key issues of dynamic simulation methods is calculation of collision impulse between rigid bodies. The correlation between contact force and relative normal acceleration could be expressed as a linear programming problem [02], which permits calculation of the collision impulse between rigid bodies colliding at multiple points. Furthermore the relation between collision impulse and relative normal velocity could be also expressed as a linear complementary problem. A simple and fast algorithm for calculating contact force with friction by formulating the relation between force and relative acceleration as a linear complementary problem was equally demonstrated [03], and this model was based on Dantzig’s algorithm (solving the linear complementary problem). Baraff’s algorithm has achieved great performance for real-time and interactive simulation of two-dimensional mechanisms with contact force, friction force and collision impulse, although friction impulse at collision was not completely covered in such a model.
for more ::->
\
http://citeseerx.ist.psu.edu/viewdoc/dow...1&type=pdf
[attachment=6574]
Nanosystem Design with Dynamic Collision Detection for Autonomous Nanorobot Motion Control using Neural Networks
INTRODUCTION
The starting point of nanotechnology to achieve the main goal of building nanoscale systems is the development of autonomous molecular machine systems. The presented paper describes the design and simulation of autonomous multi-robot teams operating at atomic scales with distinct assembly tasks. Teams must cooperate with each other in order to achieve a productive result in assembling biomolecules into larger biomolecules. These biomolecules will be delivered to “organs†(into a set of predefined organ inlets), and such deliveries must also be coordinated in time.
Building patterns and manipulating atoms with the use of Scanning Probe Microscopes (SPM) as in Atomic Force Microscopy and Scanning Tunneling Microscopy [19] is a promising approach for the construction of nanoelectromechanical systems (NEMS) with 3D precision at up to 0.01 nm resolution. However, these manual manipulations require much time and at present such repetitive tasks give imprecise results when performed manually on a large number of molecules. Approaches for nano-planning systems have been presented [19] as a first step towards automating 2D assembly tasks in nanorobotics, and the possible use of artificial intelligence as the appropriate means to enable some aspects of intelligent behaviour for the control of nanorobots in molecular manufacturing automation has been discussed in the nano community [08]. Theoretical work in molecular manufacturing has emphasized the need for very small and very accurate manipulators which simultaneously have a wide range of motion to enable the task of assembling molecular components [10]. More recent work in the possible automation of nanoscale manipulation has produced a fully autonomous motion manipulator system capable of performing 200,000 accurate measurements per second at the atomic scale [20].
•Â 2. NANOMEDICINE
The principal focus in medicine is going to shift from medical science to medical engineering, where the design of medically-active microscopic machines will be the consequent result of the techniques provided from human molecular structure knowledge derived during the 20th (and the beginning of the 21st) century [11]. For the feasibility of such achievements in nanomedicine [11] two primary capabilities are required: fabrication of parts and assembly of parts. Through the use of different approaches such as biotechnology, supramolecular chemistry, and scanning probes, both capabilities had been demonstrated in limited fashion as early as 1998 [11]. Despite quantum effects which impose a relative uncertainty to electron positions, such objections are resolved by recognizing that the quantum probability function of electrons in atoms tends to drop off exponentially with distance outside the atom, giving atoms a moderately sharp "edge". Even in most liquids at their boiling points, each molecule is free to move only ~0.07 nm from its average position [11]. Recent developments in the field of biomolecular computing [01] have demonstrated positively the feasibility of processing logic tasks by bio-computers [14], which is a promising first step to enable future nanoprocessors with increasing complexity, and nanoscale information storage and data processing capacity, which could be considered as an indispensable component of a real autonomous nanosystem. Other advances in the sense of building biosensors [25] and nano-kinetic devices [24] have advanced recently too, which could be considered as well a prerequisite for making nano-automation feasible and enabling nanorobotics control and locomotion. Many classical objections to the feasibility of nanotechnology, such as quantum mechanics, thermal motions and friction, have already been considered and resolved [10]. The presented nanorobot will be required to perform a pre-established set of tasks in the human body as is a ribosome, which is a natural molecular machine system [11].
•Â 3. PROPOSED DESIGN
A multi-robot molecular machine system could be described as a system to perform molecular manufacturing at the atomic scale, whose constituent entities are capable of cooperating collectively. Three main design approaches for nano manipulation in the liquid or air environments are: robotic arm, Stewart platform and a five-strut crank model. For our experiments we chose nano-manipulation in a liquid environment, which is most relevant within the presented application in nanomedicine. It was demonstrated that computation is relatively cheap for macroscale robotic actuators while arm motion is relatively cheap for nanoscale robotic actuators. Thus the moment-by-moment computer control of arm trajectories is the appropriate paradigm for macroscale robots, but not for nanoscale robots [11]. For nanoscale robots, the appropriate manipulator control is often trajectory trial and error, also known as sensor based motion control [16].
Virtual Environment
Virtual Reality was used for the nanorobot design where the use of macro- and microrobotic concepts is considered a practical approach once the theoretical and practical assumptions here have focused on its domain of application. The design should be robust enough to operate in a complex environment with movement in six-degrees-of-freedom. Nanoscale object manipulation systems have been applied with the use of computer graphics for teleoperation. The requirements for such systems have been clearly established [23]. A starting point for our hypotheses and experiments was to consider the robot design derived from biological models and comprised of some basic nanoscale components such as molecular sorting rotors and a telescoping manipulator (robot arm) [10]. The robot design adopted concepts provided from underwater robotics [27] keeping in mind however the kinetics assumptions that the nanorobot lives in a world of viscosity, where friction, adhesion, and viscous forces are paramount and gravitational forces are of little or no importance [11]. The obstacles will be located in unknown positions (figure 1). The delivery positions that represent organ inlets requiring proteins to be injected are located in a well-known position for the nanorobot teams if these organ inlets are (or are not) scheduled for injection at time t, they will change their colours, indicating the opening or closing of the team A (blue nanorobots) and B’s (yellow nanorobots) delivery orifice, which will indicate for the agents if they could perform their delivery in the correct order (figure 2). The trajectories and positions of each molecule are generated randomly and each molecule also has a probabilistic motion acceleration. The nanorobot navigation uses plane surfaces (three fins total) and bi-directional propellers, which are comprised of two simultaneously counter-rotating screw drives for the propulsion [11]. Considering the liquid environment, a sonar approach seems to be the most appropriate choice of sensor device for nanorobots in nanomedicine [05], thereby for navigational purposes the blue cones shown in figure 3 represent regions that the robot’s sonar can “hearâ€. Scientific visualization techniques permit rapid and precise geometric analysis for a sonar classification system [04]. The nanorobot sensors report collisions and identify when an encountered object is an obstacle to be avoided or a molecule to be caught. While some molecules are being captured (figure 3), other molecules will be assembled internally by the robot arm.
3.2 Physically Based Simulation
The study of non-penetrating rigid bodies in virtual reality for dynamic constrained simulation is a field of research in computer graphics that has an enormous impact for physically based simulation and a large range of works in this field have achieved good results. Particularly in calculating motions of many objects that move under changing constraints and frequently make collisions, one of the key issues of dynamic simulation methods is calculation of collision impulse between rigid bodies. The correlation between contact force and relative normal acceleration could be expressed as a linear programming problem [02], which permits calculation of the collision impulse between rigid bodies colliding at multiple points. Furthermore the relation between collision impulse and relative normal velocity could be also expressed as a linear complementary problem. A simple and fast algorithm for calculating contact force with friction by formulating the relation between force and relative acceleration as a linear complementary problem was equally demonstrated [03], and this model was based on Dantzig’s algorithm (solving the linear complementary problem). Baraff’s algorithm has achieved great performance for real-time and interactive simulation of two-dimensional mechanisms with contact force, friction force and collision impulse, although friction impulse at collision was not completely covered in such a model.
for more ::->
\
http://citeseerx.ist.psu.edu/viewdoc/dow...1&type=pdf
