Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
    Guy Corbaz
    @guycorbaz
    This is the gribot community instant messaging chatting tool
    afvillamil231
    @afvillamil231
    Hey Guy How are you? I couldn't answer your email yet. I'll read more about the project and the tasks later and then I will choose one!
    Thank you for adding me
    Corbaz
    @gcorbaz_twitter
    Fine. By the way, the project organization might certainly needs some improvements: do not hesitate if you have ideas.
    afvillamil231
    @afvillamil231

    Good night all. So I have been reading about the project and the progress you have made. First thank you for working on that and make it possible. Also, I have some questions about the project itself.

    Documentation

    So I checked the documentation on Git and the pdf. Also noticed that you’re changing to Latex. So I wanted to know if you have made big changes to the information and if is it possible to have access to the new one (or to the link of the Latex project).

    Specifications

    To better understand the robot and the project itself I wanted to ask you some questions. First, with respect to the specifications, I wanted to know if you have some references where I can find more information about the explanation of some of them. Like for example, why a working width of 60 cm, a ground clearance of 10cm or a travel speed of 1.8km/h. Second, should we specify what a cut grass means? For example if there is a maximum size of grass to say ok it is cut, or a percentage of the 2 hectares that has to be cut to set the task as completed. What I want to say is if there is a way out there to know that we are actually cutting the grass as we want. This can help us to be sure if our robot will be a 100% stand-alone entity or if it will require, as many other solutions out there, an additional whisper sniper or something else to finish the cut.

    Thirdly, with respect to the future of the project, It will be sold to farmers once finished? Or is it made in a way that everybody can manufacture it, program it, assemble it and deploy it in his own house? It will be important in terms of the final deliverables: an executable program to be loaded to a computer plus 3D files to print your own robot, a plug and play machine or a list of steps to build the robot. Also, that will help us to know to whom we are targeting the project: someone who knows how to weld a entire pcb, someone who knows how to use linux and program a computer or just someone who can use other opensource or not expensive tools like arduino.

    My actual work

    Now in the company where I am doing my internship, I am working with ROS and canopen. My job for the moment is to create packages and configure the ros_canopen existing ones to develop applications with a robot. So I think I can also help with that part. However, I am not sure if there is someone else in the project and on what are you guys working. I feel like before start doing things we could begin by designing and setting the entire device. Like the mechanical and electronic layer designs and the required calculations first before start programming. But I’m not really sure, I have never worked in a real project, only academic ones, so I don’t know how is it. Also, I don’t know how do you like to work or if you have a planning or a structure.

    The end

    So thank you very much again and I am sorry for the long message and the amount of questions. I hope I won’t take much of your time. Finally, I see there is a part of the documentation in french. I also speak french so let me know if you want to switch for the conversation between us. Have a good night!

    Guy Corbaz
    @guycorbaz
    Hello afvillamil123,
    First of all, thank you for your questions and comment. I will try to answer all of them, but do not hesitate to ask if there are unclear points.
    Guy Corbaz
    @guycorbaz
    The first doc have been written in Docbook, in french, because Docbook is producing nice documents. I started to translate them in english to allow more people to read them. After having discussed with some members of the project, I decided to rewrite them with LaTeX. Docbook is a free standard, but most publishing software are not free of charge. On the other hand, most LaTeX editing tools are free and easier to use. I just started to rewrite everything, and to finish the translation. I hope to be finished by the end of the weekend.
    Neither the doc nor the project is very well organized: I'll take that opportunity to define clear deliverable in order to better organize the project and the doc.
    I did not document most of the references: they are on hand written paper on my desk ;-). However, I will add them in the new written documentation.
    I chooses a working width of 60cm to allow the to move between trees and between wine grapes lines.
    Guy Corbaz
    @guycorbaz
    The ground clearance of 10cm comes from my observation in the field: agricultural parcels are not always very flat and might have holes. I addition to that, graas is higher than lawn. However, one the the project participant is working on a variable height system: this might allow to adjust the ground clearance and the height of the tools.
    A travel speed of 1.8km/h comes from 0.5m/s. When I made the first estimates, I c ome to that number as it is fast enough for what I wanted to do and limit the power and energy carried onboard. However, it is not carved in stone and can be modified if necessary.
    Guy Corbaz
    @guycorbaz
    You are righ
    You are right when you say what "cutting grass" means.The idea is that the robot start at one side of the parcel, continuously cut grass, when batteries are low returns to the base station. After loading, come back where it stop cutting in order to continue. while the grass is cut in one part of the plot, it grows back elsewhere. When the robot is finished, it mift have to start again at the begining. I would like the robot to completely cut grass, without having to manually come back with other tools.
    Guy Corbaz
    @guycorbaz
    My goal is to provide all all plans and software for free to anyone who wants to build the robot. However, I would also like to sell finished robot, or parts, or kits. The goal behind that is to allow each farmer to get a robot, even those who don't have too much money.
    At the beginning, I started to work on the software: I used ROS and simulated in Gazebo. However, a simulated robot is not "sexy" enough and not very useful for a farmer. Therefore, I paused the software development to start to work on the mechanical part. AQs I'm not a mechanical engineer, It takes some time.
    Guy Corbaz
    @guycorbaz
    We are presently 6 people working on that robot: 2 are not very active and 2 are new. That's the reason why I'm doing some advertisement for that project.
    In parallel, while discussing, We came to the idea that we could generalize this robot into an agricultural robotic platform. This could be done in designing a tool holder (or using one of the market if that exists in that size) in order to generalize the usage of the system. In addition to cutting gras, we could also weeding crops, etc.
    Guy Corbaz
    @guycorbaz
    As I mentioned above, the project and doc are not well organized, and I'm also working on improving that: I'm open to every ides/methodologies, etc. Managing an open source project based on voluntary works is rather different from the management of a project in a company: milestones can not be imposed, which means that tasks should be as independent of each other as possible. I'm basing the project methodology on "Product Design and Development" from Karl T. Ulrich and Steven D. Eppinger.
    If you have any question, please do not hesitate. By the way, if you are more comfortable in french, I'm a native french speaker. Otherwise, english is ok.
    If you know people who might be interested, do not hesitate to make some advertisement ;-): the more we are, the more fun we have (plus on est de fous, plus on ris)
    afvillamil231
    @afvillamil231
    Thank you Guy!! ok ok I think everything is clearer now. I'll wait for the document or for the LaTex link, if I can have it, to take a look or even suggest if possible. also, I can help you with the mechanical and/or electronic development If you want. I am a Mechanical and Electornic engineer from the University of Los Andes in Colombia, so I could have a useful background for the project. Meanwhile I will read about other similar projects and collect some interesting references. Thank you and looking forward to start being useful for you! bon weekend
    Guy Corbaz
    @guycorbaz
    Hello afvillamil123, I discovered a free docbook editor, https://www.xmlmind.com/xmleditor/docbook_editor.html. Using it avoid to translate everything in LaTeX. On the other hand, docbook is generating by default much nicer documents that LaTeX. If you don't mind, I would continue to use it.
    I Completely rewrote the requirement so that everyone will better understand where some numbers are comming from. The file is in https://github.com/gribot-robotics/documentation/tree/master/requirements/out/pdf. I started to work on the specification as well, but I'm not finished. Maybe could you have a look at the requirement first and tell me what you thing ?
    I'm looking to collaborate with you as it seems that you have strong experience :-) On my side, I'm an electronic engineer of EPFL Lausanne.
    ThibRbt
    @TBTRBT
    Hello the Gribot Team ! I just finished the French Robotic Cup last week with my scholar team. We decided to drive our robot by brushless motors, Odrive controller, raspberry pi and arduino mega board. The result of our performance was not as good as expected because of the discovery of the odrive controller and the total development of the motion control (odometry et servoing). Now we are stronger for the next episode ! :). Seriously, we found a real interest in this particular Odrive controller (https://odriverobotics.com/), it made for brushless motor control with encoder mount on the motor shaft, and asserv (close loop controlled) the motor by itself. The git and doc is open source but the board cost 120€ at least. That's just a proposal...In any case, I'm in to be part of the project and maybe some other members of my scholar team will too. About CAD dev, I understand that I need to get solidworks 2018 in order to design compatible parts, so I will install it asap. I will also be glad to work on programming the rolling base (pretty begging CAD see on the github!). Let's continue, hope to earing you forward, thanks. Thibault
    Corbaz
    @gcorbaz_twitter
    Hello ThibRb, I made light tests with KYDBL 2430-1E brushless controller: it works well and costs around 50$. However, the doc is rather light, and the activation of CANOpen requires a modification of the board (https://gribot.org/kydbl-2430-1e-brushless-controller/).
    I put the CADmodels on https://github.com/gribot-robotics/3d-model. Git is not the best tool for cad models, but it exists.
    The cad models are very « lights » the only part that is more or less ready is the transmission (motor-axe-wheel)
    Please tell your friends who are interested in gribot to send a message via the contact form of gribot.org, and to register on the web site, so that I could add them to the crm distribution list.
    Guy Corbaz
    @guycorbaz
    I had a look at odiver: it seems great. I'm just asking myself: will it support not only CAN bus, but CANOpen as well ? will it support position sensors ?