Implementation of Humanoid Robot Fingers Using Bio-metric Optimization
Realization of robotics with humanoid structure has become trendy in modern world. Robot fingers are a major component of robotic hands. Extent of humanoid nature vastly depends on the dexterous implementation of motion, sensing capabilities, design and effectiveness of implementation methodology. Here in this paper, it is discussed on the implementation of fingers in humanoid robot fingers by means of bio-metric optimization including design and structure, mathematical analysis and evaluation, implementation with consideration of physics and chemistry behind the designs, sensitivity of fingertips, programming and optimization of each fields to realize the humanoid robot finger close to human structure. The humanoid nature is essential in many fields such as sensitivity oriented medical surgeries, neutrally controlled hands for differently abled , visualized animation industry etc.
What is it really about?
Multi-fingered robot arms are essential in emphasizing the humanoid nature and performing various object handling fields such as hazardous and detailed productions, hospitals and homes. Effective manipulation skills, utilized joint motion behaviors and sensing abilities are valuable when considering building humanoid robot hands in a human-like manner. Coordinated movements of humanoid robot fingers are capable of firmly holding tasks.
Biometric optimization is the building of an object with the knowledge and enforcement of a biological system and its functions and qualities. Hence, the finger development of humanoid robot hands with the biometric optimization can be identified as an important research area in robotics.
Fingers, independently act a very important and wholesome task in the foresaid implementation. Hence considering biometrical factors related to fingers and relating them into humanoid robotics will cause in increase of efficiency and optimal building of a robot hand. The humanoid robot fingers has a wide variety of implementation scenarios considering different tasks of which they are objected to. Here in this paper, many scenarios considering the form and dimension of the robot fingers, sensitivity through sensors and sensitivity can be optimized, interphalangeal coordination and joint connectivity through different models have been taken into consideration and discussed with reference to different researches carried out in the world with the objective of realizing the target of implementation of humanoid robot fingers with biometric optimization to a nearly human fingers with excellent grasping, sensing , holding and pinching abilities.DBLP, Google Scholar, EI-Compendex, Mathematical Reviews, SCImago, and Scopus
What affects the implementation?
According to C. S. Lovchik and M. A. Diftler, “The Robonaut hand: a dexterous robot hand for space,” 1999, vol. 2, pp. 907–912, compatibility of the robot hand in the suggested working environment, design and the method of which linkages in robot fingers connect are key considerations in creation of highly operational robot hand. Flexible and more realistic motion and grasping abilities with using robotic fingers could be biometrically achieved in case of being able to develop technologically advanced hardware system. Zhe Xu and E. Todorov, “Design of a highly biomimetic anthropomorphic robotic hand towards artificial limb regeneration,” 2016, pp. 3485–3492. suggests that the conventional approach to designing and deployment of humanoid robot hands often involve automating biological fragments with connections, pivots and gimbals so as to reduce the apparently complex human equivalents while B.-H. Kim, “An Adaptive Neural Network Learning-based Solution for the Inverse Kinematics of Humanoid Fingers,” Int. J. Adv. Robot. Syst., vol. 11, no. 1, p. 3, Jan. 2014. suggests that adaptive neural network learning based solution for the inverse kinematics of humanoid robot fingers.
Where can we find it?
1) Medicine: medical surgeries require a considerable amount of depth perception, hand-eye coordination, decent practice for a successful performance . Hence implementation of humanoid robot hand with biometric optimization is ideal for establishing accuracy and benefactory coordination. G.-Z. Yang et al., “Medical robotics — Regulatory, ethical, and legal considerations for increasing levels of autonomy,” Sci. Robot., vol. 2, no. 4, p. eaam8638, Mar. 2017. suggests six levels of automation for robots in field of medicine as listed below.
a) Level 0: No autonomy — Robots being tele-operated on user’s command. Output of the robot includes the preferred choice of the surgeon.
b) Level 1: Robot assistance — Mechanical guidance being provided by the robot so as a human has complete control over the system. Operator has a continuous control over the system.
c) Level 2: Task autonomy — A human initiates a specific task and robot continues. Here, the operator has discrete control over the system.
d) Level 3: Conditional autonomy — Task strategies are listed by the system and continued with the choice of a human. Development of this level includes being able to continue a task without close oversight.
e) Level 4: High autonomy –under the supervision of a qualified medical professional or a surgeon, the robot makes medical decisions.
f) Level 5: Full autonomy — This level includes a “robot surgeon” who is able to make decisions and complete an entire medical surgery with best options and conclusions.
2) Space operations: vastly antropomorphic (i.e. having human characteristics) to a humanoid scale which provides fourteen independent degrees of freedom approaching good kinematics and mandatory strength, known as the “ the robonaut hand” .
3) Industrial operations : Motor vehicle manipulation, packaging, instrumentation, optics and photonics, machine assembly, industrial spray painting, arc welding, driving etc. uses robot hands widely, yet with lesser humanoid nature since the specific operations could be effectively done with specially created robot arm designs
Design and Structure of Robot Fingers with Biometric Optimization
Mathematics behind implementation of robot fingers with biometric optimization
The implementation of robot finger includes apportioning parts of the phalanx depending on relative lengths and maximum angle of deviation from a linear position.
The proximal, middle and distant lengths of each phalanx are determined as l1i, l2i, l3i respectively.
The total length of the finger i is decided by,
Li = l1i + l2i + l3i (1)
It is essential to determine a proper constraint concerning length parameters.
l1i > l2i > l3i (2)
l1i < ( l2i + l3i ) (3)
Proximal phalangeal relations are satisfied by the natural human fingers are also being taken into consideration.
l12 > l13 > l11 > l14 (4)
Hence the Inter-phalangeal Joint Coordination (IJC) is achieved in humanoid robot fingers in a biometric perspective.
θ3i = λi + θ2i (5)
λi is the Inter-phalangeal Joint Coordination parameter between Proximal interphalangeal joint θ2i (PIP) and the distal interphalangeal joint θ3i (DIP).
Motion of the distal interphalangeal joint is dependent on the actuation of the Proximal interphalangeal joint of human finger , therefore considering the relationship of 950 it is possible to create a humanoid robot finger simulating closer to that of human finger motion.
Considering the area of space grasp is S in Fig. 10 can be achieved by,
S = S1 + S2 (6)
Where,
S1=0.5 l2i2 cos α sin α (7)
And,
S2=0.5 l3i2 cos γ sin γ (8)
Biomimetic Optimization for the designing of Humanoid Robot Fingers Algorithm (BOSHRF Algorithm) has been introduced using the equations (1) to (8)
Furthermore,
(1) Initialization of length parameters of phalanx.
(2) Assigning Interphalangeal joint parameter(IJC) to the finger
(3) Includes of the length parameters of middle and distal fingers are found
(4) Extracting the includes satisfying the constraints between the phalanges given by (2) and (3).
(5) Computing the area of the grasp
(6) Estimating the total length of the finger
Force control with sensor implementation
The joints of anthropomorphic robot fingers have to be implemented in such a way to make the fingertip hit the base against the base. Together, the exceeded force application between fingertips and base has to be prevented. The fingertip controlling force can be measured using a thin tissue like force sensor (Flexi Force of Nitta Corporation) . The positional force in controlling joints of above motion is in a way of a sine wave. With the use of this technology, the delicateness of the motion and force could be retained in the humanoid robot fingers. Hence the biometrical optimization can be idolized in the fingertip motion and functionality.
Performances
In an implementation of a humanoid robot finger, performance has been measured by letting the robot hand’s thumb and other fingers hold a business card and a pen, which are both sensitive and dimension different objects. The development of the force sensors at the fingertips and motion control sensors at phalangeal inter joints has caused a successful performance in motion to hold and hold preservation of the objects.
According to , sensitivity performance has been recorded in a loop of steps which is as below
Hence the routine of even a uniform easy task has been divided into furthermore minute sub parts in order to enhance the sensitivity as well as to gain a humanoid nature and feel.
Chemistry behind optimized Robot fingers
After calculating forces exerted on different objects and the materials of which the objects are created, the next movement in implementation process includes selection of materials for the proposed design. This depends on the activity of which the robot hand is supposed to engage in. The robonaut hand designed meets the harsh conditions in space, vacuums as well as to prevent contaminations which could occur in between other space systems.
The hands of the robots has to be manufactured with more malleable (i.e. soft and flexible) material rather than sheet metal. As discussed using the equations .
The most suitable material for injection molding of robot fingers is Polypropylene with the qualities of the polymer such as availability, cost, ease of use and strength. Outstanding resistance to concentrated alcohols, acids, bases and mineral oils, higher hear resistance with a melting point of 170 °C, tensile strength of 4500 psi and a density of 938.348 kg/m3 are some features of polypropylene which has encouraged the frequent use in robot hand manufacturing process.
Software requirements
Software requirements for the development of robotic arm AVR Studio 4.18, WinAVR, Sinaprog 2.0, MATLAB v7.6. Programming the microcontroller was done using AVR Studio 4 and Sinaprog 2.0 has been used to dump the HEX file on microcontroller. WinAVR was used in backend for compiler support. Software signal processing and actuation of the servo motors has been done using MATLAB v7.6. Signal processing is also done to minimize the noise received from the motors. Interfacing between MCs and computers has been controlled by MATLAB as well and done by using serial communication.
Future Improvements
The exponential growth of computational power causes the efficiency of the creations to grow with it. In case, the flexibility and close to human nature of the humanoid robot fingers are achieved by the biometric optimization of the humanoid robot fingers. The tactile nature of humanoid robot fingers implemented of a dexterous hand by Shadow cooperation . Exact modelling of phalangeal skeleton, optimization of controlling algorithms of the fingers, designing of substitutes of rubber tendons for inter-phalangeal joints, optimized signal processing algorithms can be introduced.
So at last,
The humanoid robot hands are a major peripheral in overall humanoid robotics both visualizing and characterizing the performance of human hands. Fingers being the major components in hands, carrying out the structure and gestures of fingers with effective length design and sensing abilities are crucial. Selection of materials, considering weights of materials used in implementation, determining effective lengths and positioning of finger units in order to instigate beneficial grasping abilities is important.
Usage of sensors in detecting and controlling the motion for exceeding the desired value is a great achievement in humanizing the robotic fingers with biometric optimization. Studying about muscle flexing and attachment of tissues to each finger and the benefit of sensitivity applied by skin are biological considerations which had been taken into account in such formation.
Further, through this the conclusion of the fact that the realization of the 100% human-like movements of humanoid robot fingers cannot be made due to reasons such as cost dependencies, natural synthesis and genetically hereditary factors which differentiates the relative lengths of the fingers which cause in different abilities of picking, holding and sensations. Even though a considerable imitation and simulations can be implemented taking biometric optimization factors as discussed in above. The other major conclusion of this independent study is that the controlled movements of humanoid robot fingers can be successfully achieved using the proposed interphalangeal joint formulation. There, through the exemplary simulations using bending movements of humanoid robot fingers, the effectiveness of IJC formulation can be proved.
