Mobile Robotics

Mobile Robotics

  • Radu Tarulescu added an answer:
    Where can I find a good book for mobile robot instrumentation and measurement system?

    Hello guys, I seek for an instrumentation book which focuses on mobile robot test design and measurment systems.

    I've studied principles in instrumentation since I was undergraduate student. I am trying to design TestBeds mostly used loadcells. till know I've design and manufacture two of them and face many problems. I seek for design patterns and recommendations and how to use configuration of sensors and what are most important concerns.

    You know when you rigid a mobile robot you make it to stand and some dynamics are changed so we seek for that?

    I hope I can clear and eluminate what I think

    Radu Tarulescu · Universitatea Transilvania Brasov

    You can download many books in Robotics field from the link:

    http://www.robotee.com/index.php/download-free-robotic-e-books/

  • Anshul Joshi added an answer:
    Is it possible to control Arduino motor shield from pcduino?

    This question is very specific to the hardware I'm using. Suggestions are welcome:

    I have a pcduino that I want to control a robot with. This means controlling the robot's wheels from commands sent from the algorithm running on pcduino, via motor shields. This is easy with an Arduino+Adafruit motor shield configuration since the motor shield is designed to work with Arduino; there are no hardware issues.

    When I try to use the same motor shield with pcduino, the pcduino cannot send commands to the motor shields as (I guess) the chipsets in pcduino and arduino are different. pcduino uses an Allwinner A10 while arduino has the Atmega xxxx chip. Therefore the motor shield does not respond to these commands and errors popup in the Arduino IDE when I try to upload the sketch to pcduino.

    Is it possible to make this work without resorting to designing and building a prototype circuit on my own (I'm not EE).

    Anshul Joshi · University of Utah

    I was actually successfully able to do this. The way I did this was I connected the GPIO PWM pin of pcduino as an output, and connected it to one of the pwm pins of the arduino as input, and the signal that is passed between the two determines which action needs to be taken by the Arduino

  • Abdelkrim NEMRA added an answer:
    How to find depth from a single image?

    I´m trying to build a cheap mobile robot, capable of doing SLAM. I have a raspberry pi with a raspicam and want to find out algorithms to extract depth from monocular cues and try to implement as much as I can in opengl es since the SBC is not very strong. I know there are solutions out there (like structure from motion and make3d) but wanted to know if anyone has already try something like this…

    Abdelkrim NEMRA · Ecole Militarie Polytechnique

    Hello, you can try the LSD-SLA, it is a good solution under linux. However, be sure that with only single camera you can do SLAM up to scale (the pose anbd the map are up to scale)

    a+

  • Jose Victor Nunez Nalda added an answer:
    How do you obtain the correct dynmic model of a robot manipulator - in closed form - using Lagrange equation in matlab scrip?

    Doing dynamic model of RRRR robot manipulator, using Lagrange equation:

    D(q)∗q′′+C(q,q′)∗q′+g(q)=τ , in Matlab, the issue is that, I'm getting very long-long terms, every equation is a couple of pages, I also tried to extract the inertia matrix, again very long terms are there; I reviewed my code, and try use: simplify, rewrite, combine, factor ... commands, but doesn't work. The strange thing is that, I see there are many term involving sin and cos, similar to each others, but there are staying separately.

    Any Suggestions

    Jose Victor Nunez Nalda · Universidad Politécnica de Sinaloa

    Using twists and wrenches if I recall well can be attractive up to one point. 

    I meant that the result was the same of yours!!! The we go for comparing Newton-Euler and Lagrangian-Dynamics this in half symbolic (up to the jacobians) and then numeric substitution and further computations:

    M=m_i*J_ci^T*J_ci + J_wi^T*I_i*J_wi

    was made numerically and compared with NE formulation from which I cant recall the exact expresion of torque_i.

    The main message from our experience is twofold: a) You're not the only one getting troubles from matlab simbolic simplifications for robotics and b) I havent see the full mathematical expresion of the n oridnary differential equations for a robotic system of lets say 5 dofs...

  • Zaid Alsmadi added an answer:
    Optimum collision free path
    Hello...i am trying to develop one autonomous robot that will choose an optimal collision free path for its travel from initial to final configuration.
    I think it requires genetic programming, neural n/w and fuzzy logic approaches. Anybody can suggest me a solution and some papers
    Zaid Alsmadi · University of Malaya

    hi  Nithin,

    I believe the attached article will help you to define the concept of mean free path precisely 
    good luck

  • Dada Delimustafic added an answer:
    Which potential field is suitable for obstacle avoidance using optimal control approach?

    Is potential field or navigation potential function suitable for obstacle avoidance for optimal path planning of mobile robot using optimal control approach.

    Dada Delimustafic · Authority Partners, Inc.

    Feel free to take a look at this paper:

  • Myriam El Mesbahi added an answer:
    Is anyone using V-REP for robotics simulation of differential drive robots?

    Thanks.

    Myriam El Mesbahi · Technische Universität Hamburg-Harburg

    Thank you Eric for the support, With no doubt I'll get back to your forum if I have some implementation difficulties.

    Cheers,

    Myriam

  • Vijay Bhaskar Semwal added an answer:
    Can a mobile robot control its navigation through automatic gain controller in pid?

    i develop a manual gain pid controller for mobile robot but i want the kp,ki,kd gain should control by its own and it interact with environment factor and give stable navigational motion, how i designn this in simulink ?

  • Badr Bougattaya added an answer:
    Can anyone tip me on some articles to read for my paper on mobile robots?

    I'm doing a paper on mobile robots for my school, it's a litterature review on the existing technologies, some history and future perspective, and I need some articles to read, I was hoping to get some help finding good ones,

    Thank you in advance,

    Badr Bougattaya · Ecole Centrale Paris

    Thank you everybody for your responses, it was very helpful, I have handed my work now, and I  think I did well thanks to you again.

  • Mojtaba Karimi added an answer:
    How can we control orientation and position of a mobile robot?

    orientation and position control through matlab.

    Mojtaba Karimi · Amirkabir University of Technology

    Hi,

    if you have omni-directional moving system then you can control both of position and orientation simultaneously. in omni-directional mobile robots you can access to control of velocity in x,y and theta independently. 

    however, for feedback control of position and orientation you can use accelerometer and gyroscope data fusion with odometery.

    are u familiar with an arduino? some of arduino controller like DUE board can integrate with MATLAB and simulink. you can go to the arduino web page and search about MATLAB supported board for your project.

  • Basudev Mahapatra added an answer:
    How to extend Temporal Logic Motion Planning for mobile robots(Nao)?

    How can you apply suitable optimal path while you applying temporal logic motion planning on any mobile robot i.e. Nao. ?

    Basudev Mahapatra · National Institute of Technology Rourkela

    logic temporal follow is i think some path control being done through their dynamics analysis and their DOF. is being first study and then go for controller design .

  • Felipe Nascimento Martins added an answer:
    How can one linearly control unicycle mobile robot through pid control?

    As per path control, I design simulink function block for unicycle mobile robot, but the gain control of pid for desired position gives some unbalanced movement of robot. Can anyone suggest how I can control the movement so accurately for unicycle mobile robot.

    Felipe Nascimento Martins · Instituto Federal de Educação, Ciência e Tecnologia do Espírito Santo (IFES)

    Hi,

    You can control the unicycle using PID controllers. I suggest you check out the material of the Coursera course called Control of Mobile Robots. you will find a nice explanation and examples.

    But, if you need better performance you will want to use a nonlinear controller. The link bellow has a MATLAB/Simulink system for you to simulate controller for the unicycle. It includes the kinematic and the dynamic model of the robot, and some examples of nonlinear controllers. You can use the same model to test your own controllers.

  • Zol Bahri Razali added an answer:
    Could you recommend a paper on semi-autonomous convoy path planning/control based on noisy and slow readings of master vehicle relative position?
    Let's consider the situation that convoy consists of only two vehicles, leader and follower. Leader is driven by operator, follower can detect relative position of the leader (e.g. camera based - slow and noisy). Can you recommend a good paper on this issue?
    Zol Bahri Razali · Universiti Malaysia Perlis

    InsyAllah I will explain in detail soon.

  • Zol Bahri Razali added an answer:
    How to resolve bus checking error in KV 5000?

    I am using KV 5000 cpu unit for my Automated guided vehicles operation but suddenly its showing bus checking error. All coonections are ok. How to resolve this error.

    Zol Bahri Razali · Universiti Malaysia Perlis

    InsyAllah I will explain in detail soon.

  • Jerry Edward Pratt added an answer:
    Does anyone know companies selling series elastic actuators (SEA)?

    We need two commercially available SEAs for an experimental set-up. Can anyone recommend a company which sells such actuators? 

    Jerry Edward Pratt · Florida Institute for Human and Machine Cognition

    We have CAD models available for some linear SEAs at 

    http://ihmc.us/groups/sea/

    Luis Sentis and Nick Paine at University of Austin have some nice SEA designs and writeups about them.

    - Jerry

  • Paul F. Grayson added an answer:
    What are the factors that determine the capability of robots to navigate at a 45 degree gradient, roughly?
    In a hill like setting, the robot must grip, so that it will not topple down. What other factors are there to consider?
    Paul F. Grayson · American Industrial Magic, LLC

    Shock absorbing suspension will help keep the wheels in contact with the ground when traversing and irregular textured surface. Tire or track tread pattern appropriate for the material being driven over can also help. (pavement, sand, rock, gravel, grass, etc.)

  • Debasmit Das added an answer:
    Do you know any commercially well accepted robot simulator where (1) robot vision (2) robot sensing (3) robot control can all be simulated?
    .
    Debasmit Das · Purdue University

    While selecting simulators it is best to check what your application is : Its not always a good choice to choose the Rank 1 Simulator. For e.g

    1) OpenRave - Useful for Manipulation Planning etc. but documentation is poor

    2) Webots - Commercial and very easy to use but integration of modules and plugins are difficult. I think this software is best for undergraduate research where the student has to know basic C++.

    3) V-ReP - Simulation experience is very good and very GUI oriented

    4) Robotsim - Pretty useful for humanoids but the simulation experience is not that good.

    5) Gazebo - The Big Daddy out here which outperforms other simulators on everything because it can be universally integrated to everything.

    Right Now, the trend is ROS, ROS, ROS. So go for Gazebo

  • Amit Kumar Mishra added an answer:
    Are Kalman or Luenberger observers faster?

    Hi everybody,

    I'm working on a project in which I make a simulation of a mobile robot's equations of motion on MATLAB. I am using pole placement control using Acekerman formula for a MIMO system, and I tried to used Kalman filter as an observer, as well as Luenberger observer. 

    I tried to put some white noise on the sensor readings in the simulation, and it was obvious that the Kalman filter had better noise rejection characteristics than the Luenberger observer. However, for the same set of system's closed loop poles, it was obvious that the response in case of Luenberger observer was faster than that of the Kalman fitler (in terms of settling time and rise time). 

    My question is: does this have anything to do with the choice of the Q and R matrices of the Kalman filter? I mean if I tweaked these matrices somehow, can I get a faster response?

    Thank you in advance.

    Amit Kumar Mishra · Bhabha Atomic Research Centre

    The gain and BW of the KF depends on the ratio QR-1.  Higher ratio offers larger bandwidth, and faster transient response at the cost of high noise amplification.

  • Felipe Nascimento Martins added an answer:
    How can one design a PID controller ckt for unicycle differential mobile robot?

    Before I start ckt designing in simulink it should be able to obey the path planning and angular displacement that I have given to it.

    Can I design a simple robust design model in PID controller so that my robot will obey a circular path?

    Felipe Nascimento Martins · Instituto Federal de Educação, Ciência e Tecnologia do Espírito Santo (IFES)

    What is the role of the PID controller on your system? 

    You can apply lots of types of controllers. You can find some examples on the following toolbox:

  • Harik el Houssein Chouaib added an answer:
    Where can we find a localisation/odometry system for an indoor mobile robotic (GPS-denied environment)?

    We are looking for an odometry/localization system for indoor mobile system (GPS denied environment) that can be considered as ground truth (accuracy <cm/m).

    The result don't have to be real time, we just need a good accuracy in the path estimation.

    Harik el Houssein Chouaib · Université du Havre

    You can check this recent work, it's interesting, besides the authors has made their algorithm available and applicable on ROS:

  • Eric Nallon added an answer:
    Mobile Robot Olfaction
    Mobile Robot Olfaction (MRO) is the discipline that studies mobile robots with gas sensing capabilities. It is a relatively young discipline, and it is starting to attract attention in the robotics and chemical sensing communities.

    - Have you ever worked with gas sensitive mobile robots?
    - If so, which problem (gas source localization, gas distribution mapping, gas detection, etc.) did you address?
    - Can you point at any interesting reference dealing with MRO?
    - Which, in your opinion, are the most relevant applications of gas sensitive mobile robots?
    Eric Nallon · RDECOM The United States Army

    This is really an application of machine olfaction on a robot platform so you can benefit from all the previous research that has been done in that area like electronic nose concept.  For a mobile platform I would imagine size, weight, and power are important.  Semiconducting polymers either in a fluorescent design or in simple chemiresistor format are a good start.  Overall you are looking for an array of sensors with broadly overlapping response characteristics.  You gain information from these responses and create a feature vector which is used for subsequent pattern recognition and machine learning algorithms.  You would train your sensor array on training data and attempt to predict unknown chemical analytes.  Look into cross-reactive arrays and electronic nose, a good start is Nathan Lewis at CalTech.  It is an interesting area at the moment due to all the machine learning tools available.

  • Martin Schünemann added an answer:
    How does one include wheel slipping when modeling a car-like mobile robot using Lagrangian Mechanics?

    I am modeling the dynamics of the car-like mobile (Simplified to a bicycle model) robot using Lagrangian mechanics but so far far most literature assumes a nonholonomic constraints of no slipping. The robot is meant to move at high speed.

    Martin Schünemann · Otto-von-Guericke-Universität Magdeburg

    I would also suggest a standard tire model as Christian Oertel already did. You will need the wheel's rotational speeds and the velocity of the wheel's axle. As they are not linked over the wheel radius, you need to add the rotational dynamics of each wheel, which will increase the number of states of your model. There are a lot of books and publications on that.

    If you want to use numerical simulation of the robot dynamics, you should add some dynamics to the tire forces. Otherwise your forces and finally the acceleration of the robot are calculated directly from velocities via slip and will make your model more stiffly. The result is an oscillation of your forces, which is actually not there in reality.

  • Matin Macktoobian added an answer:
    What are the recent developments in sound localization when applied to mobile robots?
    I'm currently working on path planning and sound localization. Looking for the major contributions so far in this area.
    Matin Macktoobian · Khaje Nasir Toosi University of Technology

    Binaural sounds localization might be one the most recent one, as you can check it out bu below link...

    Of course, I recommend you to investigate on the territory of the speech recognition, since a multitude of the accomplishments in that area could be applied in various fields and applications, just like localization. 

  • Fredrick Ishengoma added an answer:
    Help me
    Frnds plz help me for downloaing papers from IEEE site. I want to present a IEEE paper as part of my studies. Unfortunatly i hav no account in their site and a sign up will cost a lot of money. In this situation it is not possible for sign up. So plz hlp me.
    Fredrick Ishengoma · Dodoma University

    Some of the IEEE journals have open access policy. look for those

  • Basudev Mahapatra added an answer:
    What is the difference between control of wheeled mobile robots (WMR) and cars and trucks?
    Why does the control of wheeled mobile robots (WMR) differ from the control of cars and trucks? Does the lateral velocity have something to do with that?
    Basudev Mahapatra · National Institute of Technology Rourkela

    hi salah,

    mobile robot depends upon kinematics and dynamics what you design and programed it, the normal car and bus which is manually controlled by human body.

  • Ahed Albadin added an answer:
    Does anyone know an algorithm for ascending stairs with a hexapod robot (insect-like, and has a hexagonal shape)?

    I have 18 DOF mobile hexapod robot (3DOF per leg) with dynamixel smart servo motor AX-12a. The motors have load feedback and I think it can help.

    Ahed Albadin · Higher Institute for Applied Science and Technology

    Thanks

  • What is the easiest way in Matlab to achieve true parallel programming?

    I have a situation where I'm controlling my robot, which has a Kinect, via Matlab. I also have matlab code that gets depth information directly from Kinect. The robots actuators behave in synchronous fashion, i.e., the Matlab code won't return unless the action is completed. Thus, during action execution, I'm not getting depth information.

    I would like to keep getting depth info from the kinect (perception) in thread/process 1 and would like to send signals to actuators, if necessary, in another thread/process. The tricky part is that the perception and action may or may not communicate with each other, i.e., they may or may not be independent. This being said, parfor etc in Matlab is not a good option for me since this is not parallel computation, rather two parallel processes that are completely different. 

    I'm using Freenect wrapper for perception and iCreate_toolbox wrapper for actuation, in Matlab.

    Omar Andres Carmona Cortes · Dalhousie University

    Anshul, the Parallel Computing Toolbox is a very powerful tool for programming parallel applications. You can use automatic parallelization (parfor), SPMD or GPU programming. However, in your particular case where you want to program completely different processes I would say that maybe Matlab is not the right choice for you. This paper gives an overview of you can do using the Parallel Computing Toolbox: http://www.europment.org/library/2014/santorini/bypaper/COMPUTERS/COMPUTERS1-05.pdf, excepting the GPU case.  I hope the article be useful. 

  • Mohamed Heshmat added an answer:
    Simple method of ground truth for visual SLAM using mobile robots

    Dear Friends,

    I want to make practical experiments for visual SLAM, and I want a simple method for the ground truth other than using Laser Range Finder of another synchronized camera.

    Thanks  

    Mohamed Heshmat · Egypt-Japan University of Science and Technology

    Dear Basudev Mahapatra,

    Thanks a lot for your answer, but how prayer and stage software can be used for making groundtruth for the robot?

    Best regards,

    Mohamed Heshmat

  • Basudev Mahapatra added an answer:
    Can anyone help with implementing ACO using ROS or Player/Stage/Gazebo?
    Need help simulating ant colony optimization for path planning in robots using ROS or Player/Stage/Gazebo.
    Basudev Mahapatra · National Institute of Technology Rourkela

    the site of kteam robot manufacture help you to install and programing purpose

  • Anshul Joshi added an answer:
    How can I represent a building floor plan for a robot?
    To be used for localization?
    Anshul Joshi · University of Utah

    It depends whether you're planning to implement this on a real robot, or on a simulation. In general, if you want a quick and dirty way(well, not really dirty) to represent a floor map, I suggest you use a 3D (or 2D, depending on your application) occupancy matrix. In 2D, an occupied element represents a wall. In 3D blocks of occupied cells represent objects. I might also have a (Matlab) code I wrote long time back, for representing maps for Kalman filter simulations.

About Mobile Robotics

A group for mobile robotics enthusiasts

Topic Followers (2,203) See all