Limbless Locomotion Robot - Coursework Example

LIMBLESS LOCOMOTION ROBOT When human race has been faced with difficulties in solving problems beyond their available technical resources, and has no basis or foundation to find a solution for the problem, engineering concepts must search for a solution developing innovative frameworks to exceed these limitations and go beyond any imaginable capabilities. This project “Serpentine locomotion articulated chain: Python” is a self-contained robot that is to evaluate the actions of the platform being able of serpentine movements. The action are assessed in a modular chain mechanical design, based on a master/slave computer architecture. Keywords: Articulated chains, snakelike robots, master-slave architecture INTRODUCTION Most of the terrestrial mobile robotic systems developed until now uses wheels or limbs as their locomotion elements. The interesting part of working with articulated chains is the ability to move through narrow paths where wheeled or legged platforms can develop some problems, opening applications on mobile robotics. Terrorist attacks and earthquakes show the evidence of the technical failure of the tools used to attend these catastrophes. In these situations, the human loss and endangerment of lives are inevitable. To save the victims of the tragedy, we have to add some necessary tools for subsequent collapses or fires. One of the most critical applications of articulated chains could be the victim’s rescue in this type of situation, where the process is accelerated, and the number of persons at risk can be diminished. The other application of a snake robot is the exploration of pipelines for maintenance. Pipelines are paths where a platform with wheels would fail while moving, places where an articulated chain will succeed. This project develops an articulated chain robot with the ability to move using serpentine locomotion. It is not expected that the platform could explore, we only intend to build a platform for future investigations. The article will present the state of the art of the project, the description, and the development of Python, as well as the results and the conclusions drawn from this work. STATE OF THE ART The work on articulated chains with serpentine locomotion began in the early 1970s. In 1972, Hirose and Morishima (1997) focused their work on platforms that could perform lateral undulation, and later they developed a series of wheeled coupled-mobility device to improve the movement of the articulated chain. They termed the devices as Active Cord Mechanisms or ACMs. Shan et al. (1997) also worked on articulated chains. They focused on work mainly in obstacles accommodation: they used the barriers to help propel the structure and not only to avoid them. They accomplished this by using a concertina mode that required a lot of space. In 1995, an articulated chain called Orochi was developed by the Japanese company, NEC, Orochi was designed mainly to search for earthquakes and natural disaster survivors making this process safer for rescuers. The device, Orochi, used an active universal joint derived from a Hooke’s joint developed by Ikeda and Takanashi (1997). Nilsson and Ojala (1997), working on the PRAIA project, at the Swedish Institute of Computer Science, developed a new serpentine universal link. This gave the robot they developed the ability to perform locomotion modes that incorporate rolling movements. The platform would be able to “hug” a tree and then roll all the way up the tree. Paap et al. (1997) and their group at GMD-SNAKE (1999) in Germany developed a snakelike device, which improves real-time control. The curvature structure required is created by the uses sensors combined with a short section of cable windings. Figure 1 Mechanical structure of ANA II. LS Robotics built a Kaa (Desai et al. 1997), which is a snakelike machine. It was created to move through support structures and a network of pipes. This was one of the first prototypes of a self-contained unit. It uses RC-servos as actuators; Kaa has a straight-line motion on flat surfaces. Dr. Gavin Miller of Interval Research Corporation has done some independent work in this area. The work began in 1987 and culminated with the prototype S5 in 1999 with a great resemblance to real snakes (Miller 2000). These are some of the developments found in the field of snakelike robots. These were the projects that were taken into consideration when developing Python for the purpose of this work. DESCRIPTION The final result of the project is the platform of Python (see Figure 1). This robot is an articulated chain of five interconnected modular segments. The platform is controlled by a microcontroller circuit master that has a serial communication (I2C) with the micro controlled circuits in each of the segments, these are called slaves. These modules are in charge of generating movement sequence and controlling the actuators’ positions, which propel each segment of the chain. The master indicates each slave’s sequence that must perform to produce the serpentine locomotion. The serpentine locomotion implemented in Python is the rectilinear mode with the abilities to make turns and manipulate the forward and backward direction of the platform. The development of the project is described in the following section. It is divided into three parts: the description of the mechanical platform, the electronic implementation, and the software of the project. DEVELOPMENT Python is a well-built platform that allows one to control the position of each segment in its vertical and horizontal axis. This is possible because of the mechanical structure of each segment built with two actuators (servomotors). The development has given the platform a freedom, which enables it to control and supervise the motion from micro controlled modules in each segment of the chain. To achieve this goal, the modules were built so that the execution of the motion(backward and forward) sequence does not depend on an external computer system. But depends on the information and communication shared between links. A general description of the master/slave computer modular implementation followed by the sequence description is described below (see Figure 2). MASTER/SLAVE IMPLEMENTATION Batteries POWER Serial 12C MECHANICAL LOCAL CONTROL Communication MAIN CONTROL PLATFORM UNIT (LCU) UNIT (MCU) 11 Acrylic Ribs Slave Module Master Module 10 Acrylic Joints Microcontroller Microcontroller 10 Servos PIC16F873A PIC16F877A Circuit Boards Operator Screws, Nuts Remote Control Wiring Batteries Figure 2 Block diagram of the master/slave implementation. Serpentine locomotion articulated chain: ANA II Wiring (Data, Microcontroller Power) Wiring (Servo System) Microcontrolled Unit Servo mounted in joint (link) Rib Rib Rib Figure 3 Lateral view of ANA II’s segment. The mechanical platform of Python- The segment structure was made in acrylic (see Figure 3).There is a head segment where the master and the reception module are placed. The segments are numbered from 1 to 5, one being the head and five the tail. The five slave micro-controlled units are interconnected by four wires, the serial communication I2C (SDA, SCL), and the alimentation (5V and ground). Python has a total length of 75 cm, a weight of 994 g, 10 degrees of freedom (two per segment), and five segments. Python was constructed and designed based on the modular concept. Its advantages are; easy reproduction, assembly, and maintenance. The platform’s modular design allows one to alter the control system of each component, giving the robot the flexibility to expand its functionalities without any changes to the robots hardware. Moreover, the platform can be increased or reduced its size without substantial changes in its architecture. Each segment consists of two acrylic “ribs”, two servos mounted on acrylic links, a slave or local micro controlled unit, wiring, screws, and nuts. Each component has two degrees of freedom, on the horizontal and vertical axis. Electronic implementation of Python- The system is constituted by a master unit and five slave units. Each slave unit in the system is in charge of the position manipulation and control of the component or segment’s servos. The operation of the slave units is independent of each other, and the master unit is in charge of conducting the chain’s performance. The slave units have the capability to handle three contact sensors and have a free port on the microcontroller for any application. The master unit can control a contact sensor, a master reset, different system configuration with a 3-bits dip-switch, a servo control, and a free Port on the microcontroller for any application. Finally, the teleoperation system consists of two parts: the transmission and reception module. The teleoperation module is a commercial one. The reception module installed on the head receives two impulse signals, which comes from the transmitter. The master unit processes these signals to execute the correspondent sequence. Control software of motion sequence of Python- Derived from the mathematical studies developed on serpentine locomotion sequence, it can be observed that the movement characteristics that have to be taken into consideration to evaluate the motion of the platform are a magnitude, frequency, and phase (Dowling 1997). Magnitude is the amplitude of the actuator’s movement. Frequency is the time an actuator takes to change from one position to another. Phase is the actuator’s initial position. These three variables are in constant interaction and affect directly the behavior of the platform The master and slave flowcharts are described later. START Initial configuration (Ports and variables setup) Movement Sequence Phase and frequency variables configuration MSSP Serial Module Configuration MAIN LOOP Remote Control Sequence Figure 4 Master module principle flowchart. Figure 4 shows the master module principle flowchart. First, the initial configuration is done. Afterward, it enters a loop where it waits for the transmission module to give a signal. When the information is received, the master divides the message and sends the sequence to its slaves. The subroutine dedicated to analyzing the specific action sent by the operator and the subroutine in charge of sending the packets containing information into the slaves is described in the flowcharts of Figure 5. Remote control Sequence MAIN - LOOP Polling RE1=Push Push_counter +1 REO RE1 REO=Enter Push Enter_ Enter_counter +1 Counter = no counter = no 0 ? 2 ? si si EXECUTE Stop Push_counter =0 Enter_counter =0 EXECUTE MAIN LOOP Push_ Push_ Push_ Push_ Counter = Counter = Counter = Counter = 1 ? 2 ? 3 ? 4 ? TX_Forward TX_Backward TX_Left TX_Right Push_Counter = 0 Enter_Counter = 0 Figure 5 The subroutine in charge of analyzing the action sent by the operator and the subroutine in charge of sending the information to the slaves. The serial communication is done by the interruption. The master unit in the system indicates when the movement of the robot should start, and the master unit does this by interrupting the operation of the slave so that it could be attended. The master unit sends five-byte streams to each slave unit. Each byte-streams has 5 bytes, and each byte has 8 bits. The five bytes sent are: Direccion (address): it is the address of the slave unit.Control Estado (state control): it indicates the sequence or command of the operator chosen; forward, backward, left, or right.Limit Superior (high limit): it indicates the higher limit at which the vertical servo can move. This variable allows one to modify the magnitude of the movement. Limited Inferior (low limit): it indicates the lower limit at which the vertical servo can move. This variable allows one to modify the magnitude of the movement. Retardo (delay): it indicates the delay between the position changes of the servo. This variable allows one to modify the frequency of the movement. Offset: it indicates the initial servo’s position. This characterizes the offset of the movement. Serpentine locomotion articulated chain: ANA II START Initial Configuration (Ports–variables configuration) Timers and Interruptions Setup Servos Center Alignment MAIN LOOP Control Control Control Control state state state state 1 ? 2 ? 3 ? 4 ? Sequence Sequence Sequence Sequence Execution Execution Execution Execution – Forward – – Backward – – Left – – Right – Figure 6 Figure shows that the slave unit is always reviewing the command that it is executing. INTERRUPTION ROUTINE SSP no TMR1 TMRO no I2C = ? = ? si si si PWM PWMs Serial Duty Period Cycle Comm Signal Timer Timer RETFIE RETFIE RETFIE 1-2 ms Period 20ms V V t t PWM Signal PWM Signal Figure 7 Figure shows the organization of the interruption routine. From Figure 6, we learn that that the slave unit is always reviewing the command that it is executing at any particular time. The slave unit will let one make the necessary changes in each segment when the master unit indicates. Figure 7 shows the organization of the interruption routine so it can attend three possible interruption sources. The TMR0 in charge of the PWM’s duty cycle for each servo. The SSP in charge of the I2C communication with the master unit and the TMR1 in charge of starting the PWM signal period. Sequence implementation- After the development of the mechanical structure of Python and its micro-controlled system, the serpentine gaits or movement sequences are implemented. Rectilinear sequence- The objective of this gait is to simulate the rectilinear gait of serpents (Lissman 1997). The same signal is sent to all the segments, but with a 90◦ offset among them. The second segment has a 90◦ offset from the first segment, the third segment has an 180◦ offset from the first segment, the fourth segment has a 270◦ offset from the first segment, and the fifth segment has a 360◦ offset from the first segment(see Figure 8).These offsets are achieved by setting the variable offset or initial position sent to each segment from the master. Thus, a vertical sine wave movement can be generated along the platform. All the segments receive a triangle signal for the vertical servo’s movement sequence.In the rectilinear gait, each segment sends a triangle signal to the vertical servo and a steady center signal to all the horizontal servos. As explained before, the master sends to each servo the initial position. In this sequence the initial position of each segment is segment no. 1, center; segment no. 2, high limit; segment no. 3, center; segment no. 4, low limit; and segment no. 5, center. Different amplitudes are manipulated varying the vertical servo position from a 0◦ to an 180◦ range divided into 100 steps. This allows a resolution up to 1.8◦ per step. These angles are taken from the center located at 90◦toward the edges or mechanical limits of the servo. A rectilinear sequence with turns- The objective of this sequence is to evaluate the platform’s turning capability. To achieve this, the vertical servos will be receiving the same signal with the same offsets to maintain the platform translation, but now the horizontal servos will move. The turning characterizes the turning radio, to the right or left of a particular amplitude. RESULTS- This type of projects is a challenge because of the complexity of the mechanical arrangement and the involvement of different areas of electronics in robotics (communications, digital systems, electrical circuits, and control) that makes complicated the development of this kind of systems. The qualitative results of the platform’s forward translation are that the speed is directly proportional to the amplitude (more amplitude, faster movement). Also, the power consumption is proportional to the amplitude, therefore, proportional to the speed. The faster the platform moves, more energy is needed to supply movement for higher amplitudes. This was an expected result because if the load of the servos increases, also the demand for current will increase. Another result is that as the frequency increases so does the platform speed of translation. We were able to prove that in order to manipulate the platform’s movement direction, the variable needed to be manipulated was the phase in the chain. This was accomplished by changing the orientation of the wave moving along the chain, modifying the initial state of the segments, therefore, the phase of the wave. The results obtained by the turning implementation showed that the platform can turn in both directions, left and right, with a turning ratio of 40 cm. With these tests, the current load was increased because of the simultaneous operation of all the servos. The platform’s evaluation and behavior on different surfaces also agreed with the previous results of magnitude and speed. The platform achieved movement in all the surfaces where it was tested. On the smooth, rough, and sand surfaces the platform was able to move forward and backward without any problem. On the stony surface the rectilinear gait was not efficient because the increase of irregularities in the surface decreased the stability of the robot using this particular gait. One of the primary advantages of this kind of platforms is their capability to keep on working when one of its segments fails because the other segments will generate enough force to move still the chain. This rarely happens with wheeled or legged platforms. The three primary variables that allow one to control the serpentine movement in an articulated chain are a magnitude, frequency, and phase. Each of these variables has an influence on the outcome of the sequence and the way the platform moves. The higher the magnitude, the greater the force that each segment adds to the chain to move; thus, the speed increases as the energy needed does too. The higher the frequency, the higher the speed of the movement, but lose to friction with the ground is the limit of this variable. The phase is the variable that characterizes the gait of the platform, indicating the initial position of each segment before the execution of a sequence. The rectilinear gait is easy to develop because it can be implemented and manipulated by the three variables mentioned in the previous paragraph, in only the vertical axis of the segment. Therefore, it is more complicated to develop an undulatory gait, which is the most common serpentine gait because it combines movement in the vertical axis and the horizontal axis. Conclusion The project gives an experimental mechanical foundation (modular structure) and a well-documented robot to develop a complete robotic system that can perform advanced operations in the segmented chain field. The platform’s modular design allows one to modify the control system of each chain, giving the robot the flexibility to expand its functionalities without any changes to the robots hardware. Besides, it has the advantages of any modular design, easy reproduction, assembly, and maintenance. Total segments are limited by the servo’s torque. If the chain is too long, the middle chain servos have to be strong enough to move. This also is related to the platform’s weight, a variable that has to be always in mind. In a long-term the project can reach the stage of completely functional recognition and exploration system, pipe maintenance, search and rescue system for natural disasters, military intelligence system, etc. Acknowledgment I would like to acknowledge everyone who contributed into the successful completion of this project. I would like to thank my professor, friends and family members who offered their moral and professional support. REFERENCES Desai R, Rosenberg CJ, Jones JL. 1997. Kaa: An autonomous serpentine robot utilizes behavior control. In: Dowling K (Ed.), Limbless Locomotion: Learning to Crawl with a Snake Robot. Thesis (Doctor of Philosophy in Robotics), Carnegie Mellon University, The Robotics Institute, Pittsburgh, p. 31. Dowling K. 1997. Limbless Locomotion: Learning to Crawl with a Snake Robot. Thesis (Doctor of Philosophy in Robotics), Carnegie Mellon University, The Robotics Institute, Pittsburgh, pp. 61–6. GMD-SNAKE. 1999. Robot-snake with flexible real-time control. GMD’s Institute for Autonomous Intelligent Systems. http://ais.gmd.de/BAR/snake.html Hirose S, Morishima A. 1997. Design and control of a mobile robot with an articulated body. In: Dowling K (Ed.), Limbless Locomotion: Learning to Crawl with a Snake Robot. Thesis (Doctor of Philosophy in Robotics), Carnegie Mellon University, The Robotics Institute, Pittsburgh, pp. 21–4. Ikeda H, Takanashi N. 1997. Joint assembly moveable like a human arm US Patent 4683406. In: Dowling K (Ed.), Limbless Locomotion: Learning to Crawl with a Snake Robot. Thesis (Doctor of Philosophy in Robotics), Carnegie Mellon University, The Robotics Institute, Pittsburgh, pp. 27–8. Lissman HW. 1997. Rectilinear motion in a snake (Boa occidentalis). In: Dowling K (Ed.), Limbless Locomotion: Learning to Crawl with a Snake Robot. Thesis (Doctor of Philosophy in Robotics), Carnegie Mellon University, The Robotics Institute, Pittsburgh, pp. 17–8. Miller G. 2000. Snake robots. Private research project into snake locomotion started in 2000. http://www.snakerobots.com/ Nilsson M, Ojala J. 1997. Self-awareness in reinforcement learning of snake like robot locomotion. In: Dowling K (Ed.), Limbless Locomotion: Learning to Crawl with a Snake Robot. Thesis (Doctor of Philosophy in Robotics), Carnegie Mellon University, The Robotics Institute, Pittsburgh, p. 29. Paap KL, Dehlwisch M, Klaassen B. 1997. GMD-Snake: A semi-autonomous snake like robot. In: Dowling K (Ed.), Limbless Locomotion: Learning to Crawl with a Snake Robot. Thesis (Doctor of Philosophy in Robotics), Carnegie Mellon University, The Robotics Institute, Pittsburgh, p. 30. Shan Y et al. 1997. Design and motion planning of a mechanical snake. In: Dowling K (Ed.), Limbless Locomotion: Learning to Crawl with a Snake Robot. Thesis (Doctor of Philosophy in Robotics), Carnegie Mellon University, The Robotics Institute, Pittsburgh, pp. 26–7. Read More