2018_Team15

Project Overview  |  Proof of Concept  |  Final Design  |  Fabrication  |  Testing and Results  |  Meet the Team  |  Acknowledgements


Project Overview

Cable driven parallel robots (CDPRs) is a type of robot that utilizes cables in place of rigid components to achieve a greater range of mobility. Team SpacePack aims to redesign a CDPR for the use of human interactive simulations. With NASA Johnson Space Center as the team’s sponsor, the goal of the project is to simulate the dynamic environments that an astronaut will experience on mars. A common issue that has been encountered is maintaining tension in the cables during movement of the CDPR’s end effector (platform) within its defined workspace. By utilizing multiple sensors and correcting the feedback error of the fused sensors the team plans on integrating a controller within the CDPR’s software to maintain tension and achieve positional accuracy of the end effector. Software will be written in Matlab and will eventually be translated into Python. The team plans on providing an affordable sensor development for analysis of robots actuated by cables (ASDARAC). The team will build a small-scale, 6 degree of freedom (6-DOF) prototype as their final product. Despite the team’s agenda to create a simulator for use in astronaut training, the CDPR can also be applied within the gaming industry, the shipping industry, and warehouse production. The team will deliver its final product to NASA Johnson Space Center.

Cable driven parallel robots (CDPRs) is a type of robot that utilizes cables in place of rigid components to achieve a greater range of mobility. Team SpacePack aims to redesign a CDPR for the use of human interactive simulations. With NASA Johnson Space Center as the team’s sponsor, the goal of the project is to simulate the dynamic environments that an astronaut will experience on mars. A common issue that has been encountered is maintaining tension in the cables during movement of the CDPR’s end effector (platform) within its defined workspace. By utilizing multiple sensors and correcting the feedback error of the fused sensors the team plans on integrating a controller within the CDPR’s software to maintain tension and achieve positional accuracy of the end effector. Software will be written in Matlab and will eventually be translated into Python. The team plans on providing an affordable sensor development for analysis of robots actuated by cables (ASDARAC). The team will build a small-scale, 6 degree of freedom (6-DOF) prototype as their final product. Despite the team’s agenda to create a simulator for use in astronaut training, the CDPR can also be applied within the gaming industry, the shipping industry, and warehouse production. The team will deliver its final product to NASA Johnson Space Center.

BACK TO TOP


Proof of Concept

The proof of concept model consists of two motors attached to an end-effector or point mass. The two motors, placed directly opposite of each other, will pull the end-effector in opposite directions through a cable-pulley set up. Two separate cables will be assigned to each motor and attached onto the end-effector at opposite locations. The proof of concept is a 1 degree-of-freedom, linear design that will pull the end-effector in a side to side fashion. With this concept, the configuration of sensors to accurately provide state estimation of the end-effector will be finalized. A partial-integral-derivative controller will be applied to our proof of concept in order to assure that our end-effector can timely and accurately follow a given trajectory. Through the proof of concept, seamless communication between the cable robot and its user software will be achieved through the use of an Arduino and a laptop.

A cable driven parallel robot is considered functional when the robot is able to move its end effector within its workspace in a collision-free and tangle-free fashion. It is also crucial to maintain tension within the actuating cables of the robot to achieve an enhanced performance. By always maintaining tension in all cables, precise and accurate position and acceleration control of the robot can be achieved. A feedback PID controller will be designed as the robot’s control system. The controller relies on a multi-sensor system for state estimation and localization of the robot. The motors used to move the robot must also be able to provide a sufficient amount of torque for the seamless movement of a maximum payload.  The CDPR will help train astronauts maneuver a rover on varying celestial bodies. CDPR’s can also be applied to the shipping industry by moving a large containers and weight at shipping ports.

BACK TO TOP


Final design

Team Space Pack is designing a cable driven parallel robot (CDPR). The goal of the project is to design a tension centered controller within a 6 degree-of-freedom cable robot for maintaining tension within all cables such that the robot’s end-effector may experience various types of dynamic environments. The team is designing the cable-driven robot as a small scale prototype for NASA Johnson Space Center (JSC). NASA JSC plans to integrate a tension controller within a large scale CDPR to be used in future astronaut training programs. By providing NASA with our prototype, Team Space Pack will have completed the necessary research on how to incorporate a tension controller within a cable-driven robot. Additionally, the team will be able to provide NASA with direction towards applying tension sensors within a large scale cable robot.

The difference in design objectives between Team Space Pack’s project, Affordable Sensor Development for Analysis of Robots Actuated by Cables (ASDARAC), and competitors also developing cable robots is that the team’s goal is to simulate a specific environment onto the end-effector. A product design specification of the project is to be able to apply 4 gs of acceleration onto the end-effector. Competitors typically design cable robots with positional accuracy and speed as the main design objectives. Competitors tend to focus on the aforementioned design characteristics due to the application of their robots in logistics and manufacturing. The purpose of these robots include simply moving an object, such as a box or pallet, from one point to another. In order to perform this action, competitor CDPRs are human-controlled through the use of a joystick. Team Space Pack focuses on cable robot research and development for vehicular simulation purposes, hence, the user interface is used in competitor’s robots are not applicable to Team Space Pack’s robot.

The ASDARAC project will produce a CDPR that is autonomous and requires only an input of the desired trajectory that the end effector will trace. The trajectory can be tailored to provide a specific simulated environment. Since the ASDARAC project will produce a CDPR that does not include a gripper at the end-effector, the costs associated with developing the gripper will be avoided. Instead, a simple platform that encompasses a 1 kg weight is featured in the ASDARAC cable robot. Due to the ASDARAC project’s objective of creating simulated dynamic environments, software that enables the control of the loads that are experienced by the end-effector is developed. Finally, the CDPR will include multiple, low-cost sensing capabilities through the inclusion of three low-cost sensors as opposed to competitors who use expensive sensors that make CDPRs unaffordable.

The final prototype of the robot is a six-degree-of-freedom (6-DOF) system. The end-effector must move in the X,Y,Z planar directions, and must achieve roll,pitch, and yaw motions (rotation about center x, y, z axes). The 6-DOF CDPR cubed frame will be composed of aluminum alloy. The design consists of eight motors and eight individual cables. Motors are placed in the vertices of the cube. The end-effector is also cube-shaped.

BACK TO TOP


Fabrication

Several components of team SpacePack’s prototype were fabricated through machining, 3D printing, and woodworking.  The prototype’s base is a piece of thick plywood sawed to the necessary dimensions from a  larger piece of plywood. Several, long, 1 meter tall, T-slot aluminum tubes were cut in the Mechanical Engineering Department’s machine shop by the the team members. The newly cut metal bars were used to construct the cubic structure of the robot atop the plywood base. Additional shelf-like sections at the top and bottom of the cube are utilized to mount the sensors, motors, and microprocessors outside of the cube so that the robot’s workspace is maximized. Readily available and compatible brackets were ordered for the assembly of the prototype so that uneven assembly of the robot due to self-machined brackets was avoided. Holes in the wooden base, motor mounting brackets, and sensor mounting brackets were drilled using tools from the machine shop. In order to accurately and properly cut, drill, and mate all components of the prototype, team SpacePack created and printed official drawings for all prototype assembly components.  Drawings were handed over to the shop and included hole placement and part dimensions for ease of machining. After upgrading the prototype’s motors to those providing higher torque than the proof of concept, it was decided that metal spools were better suited for the system. The spools needed to drive the robot’s movement were designed in SolidWorks and machined using a CNC milling machine and a CNC lathe. The spools contain a small hole in their cylindrical base so that a small cable can be fed into the inner cylinder of the spool. Lastly, the endefector was 3D printed using the printer in the machine shop.

After all machined parts were created, the capstone team used carefully fitted screws and bolts  to begin assembling the prototype. First, the rectangular aluminum base of the robot was formed and bolted together using T-slot, L-piece corner brackets. Then, the top and bottom mounting regions were constructed. Once the metallic frame for the mounting was completed, wool panels were inserted and leveled on the top and bottom of the newly formed cubic workspace using mounting L-brackets. After the main structure was assembled, all of the mounts for the sensors, motors, and microcontrollers were placed. The four upper motor and sensor systems were mounted upside down on the top mounting region,. The four bottom motor and sensor systems mounted to the base of the robot are placed within the bottom mounting region.

Once the motor and sensor systems were added to the robot’s structure, swivel pulleys were mounted in all eight corners of the main frame. The motors were aligned such that they properly operate with the pulleys. About 6 feet of thin, wire cable was placed around each spool by feeding the cable through a hole placed on the spools cylindrical face and tying a knot at the end of each cable so that it may no longer fit through the feed hole. The other end of the each cable was threaded through the pulleys and then tied to the end effector so that the physical structure of the robot becomes complete. Arduinos, a data acquisition system, and a power supply are all connected to the motors and placed along the base of the robot for easy access.

BACK TO TOP


Testing and Results

Team SpacePack tested the motors, tension sensors, and encoders (motor rotation sensors) of their 6 degree of freedom (DoF) cable driven parallel robot (CDPR). The team aimed to receive feedback from both the tension sensors and encoders during movement of the robot’s actuation motors. The actuation of the motors is expected to move the end-effector.  The team programmed the motors to test whether Arduinos can be used to rotate all motors in the clockwise and counterclockwise directions at different speeds. The Arduinos used in the robot were able to achieve this feat. While the robot’s motors are rotating, the tension sensors connected to the motors’ accompanying pulley and kevlar cable system are tested to read the tension in the kevlar. While tension in the kevlar cables increased, the graph output from our tension sensors proved to give real time data according to deflection caused by cable tension. The encoders attached to the motors were then tested for functionality by rotating the motors and implementing a rotation counting code to count the number of shaft rotations. The team was able to display the number of shaft rotations on a monitor according to motor actuation. The team tested the prototype over 20 times. The prototype did not initially work as planned. The team primarily made sure all the individual components worked separately, by coding them and testing them, and then attempted to incorporated all sensors together. During prototype testing, the end effector was moved (as seen in the video) and tension data was read as seen in fig 1. The problem Team SpacePack’s prototype addresses is the feasible reception of tension feedback with low-cost material and electronics. Users of this technology are going to be manufacturing companies or aerospace companies which will have to spend less money for the similar sensing technology. NASA Johnson Space Center, the sponsor and the main proprietor of Team SpacePack’s project, were happy with the product when the team presented all components of the 6-DOF CDPR.

LINK TO VIDEO:

https://drive.google.com/open?id=1OXklaTyijLKB73wFANOjkpb_7NF7Y3NX

BACK TO TOP


Meet the Team

 

Nikhil Sidher: Nikhil is a currently a senior Mechanical Engineering student. He’s also pursuing minors in math and unmanned autonomous systems. Nikhil is interested in getting a Master’s Degree in robotics or controls engineering after graduation. A major project he’s worked on is building a fixed-wing drone with VTOL capability. Nikhil has a lot of experience coding in MATLAB and C and modeling in SolidWorks from his coursework. Nikhil is expected to graduate in December 2018.

 

 

 

Natali Salas-Espana: Natali is from Carson City, Nevada. She is currently a Mechanical Engineering (ME) student at the University of Nevada, Reno. She is also completing her minor studies in Mathematics. Additionally, Natali is a member of Shan Research Group (SRG) where she is currently in the testing phase for her project on tune-able dynamic friction within a polydimethylsiloxane (PDMS) chip. PDMS is a popular material used in prosthetics and soft robotics. Natali’s goal before graduating is to design a soft crawling robot that utilizes dynamic friction to change its direction of locomotion. Through her coursework and research, Natali has gained experienced developing code in MatLab, 3-D drafting molds in SolidWorks, performing FEA analysis on mechanical components, and handling temperature sensitive materials. After graduation, Natali plans to enroll in a Master’s program in Mechatronics and Robotics to combine her smart material research with rigid robotics in order to develop human assistive robots.

 

 

 

Austin Lopez: Austin is currently a senior studying Mechanical Engineering at the University of Nevada, Reno. Austin is working under Dr. Yan Wang where he simulates heat transfer using molecular dynamics. Additionally, Austin Lopez has a strong background in robotic development. Austin has designed and developed an autonomous drone for package delivery. Through Austin’s academic career he has developed several technical skills that are useful for engineering. Austin has learned how to code with Matlab, model 3D objects and run FEA analysis with SolidWorks, and utilize Labview for experimental measurements. After graduation Austin plans to attend graduate school to obtain a PhD in control systems and robotics.

 

 

Inrianto John A. : John was born in Merauke, Indonesia. John is currently a Mechanical Engineering student at University of Nevada. He is also pursuing minor in math and unmanned autonomous system. John has a lot of experience working with 3-D modeling on SolidWorks and data analyzing with Matlab. John is expected to graduate in May 2018.

 

 

 

 

Manuel Retana: Manuel’s hometown is San Miguel de Allende, Guanajuato, Mexico. One of Manuel’s hardest engineering projects took place when he studied abroad at University of Bristol in England. He enrolled in the graduate group design capstone class where he designed a satellite for solar imaging with nine other masters’ students. The project was sponsored by Airbus Defense and Space Corporation. Manuel has learned how to conduct research, solve problems, create models and analyze complex dynamics systems both at UNR and at his co-op tours at NASA JSC. Manuel founded both the mariachi and taekwondo clubs at the University of Nevada, Reno to share his passions outside of school. This summer Manuel conducted research at Stanford University where he combined to completely different space rovers to explore asteroids. He developed simulations in which the two robots worked together to survey the surface of an asteroid. Manuel goals are to go to graduate school and continue his co-op at NASA. In the long term, Manuel plans to become a NASA researcher in space robotics after he gets a PhD. He wants to develop robots to help astronauts and humanity get to Mars and the Moon.

 

 

 

 

 

 

 

BACK TO TOP


Acknowledgements

Team SpacePack thanks the following entities for their support and help:

-NASA JSC (Sponsor)

-Joshua Sooknanan (NASA Engineer)

-Dr. Kostas Alexis (Capstone Mentor)

-UNR Undergraduate Research Office (Project Funding)

-UNR Mechanical Engineering Department (Project Funding)

-UNR ME Machine Shop (Providing tools, basic materials, and knowledge of machining)

BACK TO TOP