與 V-rep 有關的訊息

https://web.archive.org/web/20170313212157/http://blueworkforce.com/uncategorized/merger-blue-workforce-acquired-coppelia-robotics-the-v-rep-company/

http://project.mde.tw/blog/blue-workforce-bing-gou-v-rep-gong-si.html

https://onrobot.com/en/new-acquisition-onrobot-secures-several-new-technologies-and-products

Blue Workforce A/S and Coppelia Robotics GmbH – Merged 28th December 2016.

Blue Workforce A/S , Denmark and Coppelia Robotics GmbH, Switzerland,  and has 28th December merged to pursue a shared vision on making robotics more Accessible, Applicable and Affordable.

The decision is the outcome of 3 years of intensified talks, idea sharing and mutual inspiration on how the highly innovative products and business models on both sides demonstrate the power to radically relax and improve how robotics can be conceived by both, low- and high-end users.

It became clear that a merged technology assert pool unbound, was the right way to proceed with the vision we shared. Both founders, Marc Freese from Coppelia Robotics GmbH, and Preben Hjornet from Blue Workforce A/S are experienced Roboticists with a long track records of bringing innovations into the global robot industry.

“to me it became clear, that the V-REP simulation tool was a “one of a kind” and the only one which gave me confidence that I could model behavior close enough to be valid enough to give guarantee´s on the performance shown”.  Preben continues “Once I got into direct contact with the inventor and founder of V-REP, Marc Freese, things began to move fast forward between us, on how we could merge the robotics platform of Ragnar Robotics and V-REP.

Beyond the superior simulation capability, we have proven that the two together could add even more value to us and our customers as an emulation and real-time VR tracking tool. This has internally been a central part of our test, optimization and validation processes developing the Ragnar robotics”. Marc states “ It is a move that was mend to happen, and I’m excited about joining with Blue Workforce, and the extended opportunities it brings “ .

NEW ACQUISITION: ONROBOT SECURES SEVERAL NEW TECHNOLOGIES AND PRODUCTS

The rights to the unique robot technology behind the robot called RAGNAR have been bought by the Danish company OnRobot, because the management sees obvious synergy possibilities with OnRobot's product range. 12 robot developers from the bankrupt Danish company Blue Workforce will at the same time get a new employer as OnRobot has offered them work as developers in the successful international robot company.

Shortly before Easter, negotiations between OnRobot and the curator of Blue Workforce fell into place. It happened just 10 days after Blue Workforce was filed for bankruptcy as a result of a failed funding round. The company employed 30 employees before the bankruptcy, and OnRobot now employs 12 of these.

OnRobot has taken over all the design rights for the products in Blue Workforce as well as machines, inventory, and tools. These will be integrated into the work of OnRobot's development department, which designs equipment and gripping tools for automation solutions with collaborative robot arms of all makes, also known as collaborative applications. Both small and large manufacturers use the products from OnRobot in the manufacturing industry globally. Blue Workforce, founded in 2012, has in particular delivered solutions for the food industry and agriculture.

“We saw an obvious opportunity to expand our product portfolio with some unique technologies and competencies within soft gripping and vision technologies, i.e. solutions that can handle delicate items and foods without damaging them as well as camera-based solutions for inspection on production lines. We expect very quickly to be able to create new OnRobot products with ingredients from the inventions we’ve bought. We have nine different products today, and within the next 21 months, we would like to reach 40-50 products, grippers and sensors and vision and other technologies for collaborative robotics applications. So, by the end of 2020, we expect to be ready with a complete product range of equipment for collaborative robot solutions,” says Enrico Krog Iversen, CEO of OnRobot.
 
12 skilled robot developers so far employed by Blue Workforce thus received a lot of new colleagues, mainly in Denmark, from Monday 15 April, but also in Hungary, Germany, Poland, Spain, USA, China, and Singapore. OnRobot's philosophy is that factors such as geography and logistics should not stand in the way of recruiting the world's best robot technology developers. Therefore, the development team from Blue Workforce can continue to work from a local development department in Aalborg, Denmark. The hiring of the developers has taken place in a process where OnRobot's CTO, Niels Degn, has met every robot developer. With the 12 new colleagues, OnRobot's total staff now counts just over 145 employees globally, of which 50 work at the head office in Odense, Denmark. At the end of 2019, the number of employees is expected to have grown to 175 employees. Also, OnRobot is expanding its global dealer network from 200 to 1000 retailers over the coming years.

About OnRobot

Founded in 2015, OnRobot merged with Perception Robotics and OptoForce in 2018 and also acquired Purple Robotics in the autumn of 2018. In April 2019, Blue Workforce's rights were also acquired along with 12 developers. OnRobot's product range includes a wide range of robotic equipment such as electric grippers, force and torque sensors, vacuum grippers and tool changers.

The new product mix from OnRobot makes it quick and easy to automate tasks such as packaging, quality control, material handling, machine operation, assembly, and surface finishing. The head office is in Odense, Denmark, and OnRobot also has branches in Hungary, Germany, Poland, Spain, USA, China, and Singapore. The company plans to expand further through acquisitions in the coming years.

V-rep 與 Webots 機器人模擬系統

2013 年 1 月 10 日,  V-rep 宣布開源, 以 GPL3 使用授權釋出, 允許非營利學校免費使用, 商業使用必須付費, 檔案格式未公開.

2018 年 12 月 18 日, Webots 宣布開源 (https://github.com/omichel/webots), 並以 Apache 2 使用授權釋出, 檔案沿用 VRML97 格式, 使用與延伸上幾乎沒有任何限制.

V-rep 與 Webots 使用者手冊轉入 CMSimfly

網際內容管理實習專案

V-rep

http://www.coppeliarobotics.com/helpFiles/ (Github 倉儲) 為一系列 html 檔案, 可從 V-rep 3.6.1 helpFile 檔案下載, 為了讓使用者能直接利用 CMSimfly 編輯管理此一文件資料, 請各組討論該如何將這些資料內容轉為 CMSimfly 格式的靜態與動態網頁.

靜態網頁請使用 Github Pages, 同步動態網頁則採用 Heroku 製作.

Webots

https://cyberbotics.com/doc/guide/menu (Github 倉儲) 中的 docs 目錄則有一系列的 Markdown 檔案, 目前 Webots 採用 Javascript 以動態模式, 將 Markdown 文件轉為 html, 並且透過倉儲 tag 的版次號, 讓使用者可以直接在網際以 URL 參數擷取不同版次下的技術文件.

Webots 目前對於技術文件處理的流程, 除了能夠將網際文件 (目前有英文版的 guide, reference 與 automobile 等三個文件目錄) 以分散式多人協同進行管理外, 對於之後建立不同語系技術文件, 也相對較有效率, 因為只要透過 URL 的參數指向不同語系 Markdown 所在目錄即可讓使用者選擇各語系、各版次 (透過倉儲 tag release) 的網際技術文件.

例如:

以下為 R2019a 版次, guide, reference 與 automobile 目錄下的文件.

https://www.cyberbotics.com/doc/guide/page?version=R2019a

https://www.cyberbotics.com/doc/reference/page?version=R2019a

https://www.cyberbotics.com/doc/automobile/page?version=R2019a

R2019a-rev1 版次文件則分別如下:

https://www.cyberbotics.com/doc/guide/page?version=R2019a-rev1

https://www.cyberbotics.com/doc/reference/page?version=R2019a-rev1

https://www.cyberbotics.com/doc/automobile/page?version=R2019a-rev1

至於 CMSimfly 直接讓使用者編輯 html 檔案的方式, 只要設法引入網際 Markdown Editor, 並採用上述 Webots 處理 Markdown 文件與版次的方式即可!?


參考資料

https://www.crummy.com/software/BeautifulSoup/bs4/doc/

https://pythonprogramminglanguage.com/get-all-image-links-from-webpage/

https://stackoverflow.com/questions/20132458/how-to-concatenate-two-html-file-bodies-with-beautifulsoup

https://stackoverflow.com/questions/39292881/python-beautifulsoup-replace-img-src

自行編譯 V-rep

背景環境

2018 年除了在 https://github.com/mdecourse/cd2018/issues/15 中記錄如何自製 Windows 10 64 位元系統中的可攜式 3.5.0 rev3 V-rep 外, 還嘗試自行利用 MSYS2 編譯 V-rep, 當時使用的可攜系統為 kmol2018_for_compiling_vrep.zip, 而所完成的結果為 3.4.0 rev1 serialization version 20 的 vrep_kmol.7z.

進入 2019 年, https://github.com/CoppeliaRobotics/v_rep/releases 已經釋出 3.6.0 版, 現在 KMOLab 希望探究, 相同的流程是否能夠完成最新版本 V-rep 的編譯.

當然, 在 Windows 64 位元系統中, 自行編譯 V-rep 的目的, 也是希望透過對於原始碼的研究過程中, 進一步了解 V-rep 相關模組配置的細節, 設法將此一系統的最大功能發揮出來.

首先, 看看 V-rep 3.6.0 相對於 3.4.0, 功能有哪些差異? 從 https://github.com/CoppeliaRobotics/v_rep/releases/tag/vrep-v3.6.0 來看, 隻字未提, 但是從 https://github.com/CoppeliaRobotics/v_rep/releases/tag/v3.5.0 可以知道 3.5.0 版在 2018.02.07 釋出, 因此想知道從 3.4.0 到 3.6.0 V-rep 核心程式之間的差異, 只能自行從累次的提交內容中, 自行爬梳.

接下要進行的工作, 除了設法將最新版製作成可攜之外, 就是一連串程式編譯流程中資料的探索, 目前任務尚未完成, KMOLab 必須繼續努力.

第一個完成的任務為: V-rep 3.6.0 rev1 的可攜系統 v-rep3.6.0_rev1.7z, 先試試看能否在一般的 Windows 10 64 位元系統中執行.

執行編譯

編譯 V-rep 的第一步, 是確認系統所使用的  Lua 版本, 因為整體的 V-rep 架構都是植基於 Lua 系統原始碼之上, 截至 V-rep 3.6.0, 官方仍然使用 Lua 5.1 (目前 Lua 最新版已經來到 5.3.5), 從 https://www.lua.org/source/5.1/ 可以查看 5.1.5 原始碼中的內容, 並且可從 https://www.lua.org/ftp/ 下載 Lua 5.1 的原始碼, 之後在命令列中更換目錄到 lua-5.1 之後, 執行 mingw32-make mingw, 表示要使用 MSYS2 中的 mingw32-make 執行 MAKE, 且宣告 platform 為 mingw, 完成編譯後, 可以在 src 目錄中找到 lua.exe 與 lua51.dll.

註: mingw32-make mingw 指令可以正確編譯 lua 5.1, lua 5.1.4, lua 5.3.4 與 lua 5.3.5, 唯獨無法正確編譯 lua 5.1.5, 因為這裡只需要 lua 5.1, 因此不再處理 lua 5.1.5 編譯上的問題.

由於 V-rep 系統中也使用 lua51jit.dll, 因此可以從 http://luajit.org/download.html 下載 LuaJIT-2.0.5, 進入目錄後, 以 mingw32-make 即可完成編譯.

Lua 5.1.5 除錯

前面以為可以無需處理 Lua 5.1.5 編譯問題, 一旦進入 https://github.com/CoppeliaRobotics/v_rep/blob/562bd9fa1b22b9919a9bee03df64436b21529174/config.pri, 赫然看到 V-rep 開發者在 Windows 建議使用的 Lua 版本, 竟然就是 5.1.5 版, 因此, 還是必須處理所碰到的問題.

從 http://lua-users.org/lists/lua-l/2012-08/msg00121.html 得知 Lua 5.1.5 在釋出原始碼之後, 才發現幾個錯誤, 內容記錄於 http://www.lua.org/bugs.html#5.1.5-1, 只要按照說明更新錯誤, 即可利用 mingw32-make mingw 順利編譯出隨後所需的 lua.exe, luac.exe, lua51.dll 與 liblua.a.

下載 Kmol_lua_515_535.7z

編譯 V-rep 核心

第一步必須編輯 y:\vrep360_kmol\v_rep 目錄中的 config.pri 設定檔案, 根據目前所使用的系統設定內容修改如下:

# location of boost headers:
    BOOST_INCLUDEPATH = "y:/boost_1_66_0"    # (e.g. Windows)
    #BOOST_INCLUDEPATH = "/usr/local/include"    # (e.g. MacOS)

# location of lua headers:
    LUA_INCLUDEPATH = "y:/lua-5.1.5/src"    # (e.g. Windows)
    #LUA_INCLUDEPATH = "../../../../../mingw64/include/lua5.1"    # (e.g. Windows-MSYS2)
    #LUA_INCLUDEPATH = "../../lua5_1_4_Linux26g4_64_lib/include"    # (e.g. Ubuntu)
    #LUA_INCLUDEPATH = "/usr/local/include/lua5.1"    # (e.g. MacOS)

# lua libraries to link:
    LUA_LIBS = "y:/lua-5.1.5/src/liblua.a"    # (e.g. Windows)
    #LUA_LIBS = -llua5.1    # (e.g. Windows-MSYS2)
    #LUA_LIBS = -L"../../lua5_1_4_Linux26g4_64_lib/" -llua5.1    # (e.g. Ubuntu)
    #LUA_LIBS = "/usr/local/lib/liblua5.1.dylib"    # (e.g. MacOS)

# qscintilla location:
    QSCINTILLA_DIR = "y:/QScintilla_gpl-2.10.2"    # (e.g. Windows)
    #QSCINTILLA_DIR = "../../QScintilla-commercial-2.7.2"    # (e.g. Windows-MSYS2)
    #QSCINTILLA_DIR = "../../QScintilla-commercial-2.7.2"    # (e.g. Ubuntu)
    #QSCINTILLA_DIR = "../../QScintilla-commercial-2.7.2"    # (e.g. MacOS)

# qscintilla headers:
    QSCINTILLA_INCLUDEPATH = "$${QSCINTILLA_DIR}/include" "$${QSCINTILLA_DIR}/Qt4Qt5"

# qscintilla libraries to link:
    QSCINTILLA_LIBS = "$${QSCINTILLA_DIR}/libqscintilla2_qt5d.dll.a"    # (e.g. Windows)
    #QSCINTILLA_LIBS = "$${QSCINTILLA_DIR}/release/release/libqscintilla2.dll.a"    # (e.g. Windows-MSYS2)    
    #QSCINTILLA_LIBS = "$${QSCINTILLA_DIR}/release/libqscintilla2.so"    # (e.g. Ubuntu)
    #QSCINTILLA_LIBS = "$${QSCINTILLA_DIR}/release/libqscintilla2.9.0.2.dylib"    # (e.g. MacOS)

# Make sure if a config.pri is found one level above, that it will be used instead of this one:
    exists(../config.pri) { include(../config.pri) }

接著在命令列中, 進入 y:\vrep360_kmol\v_rep 目錄, 執行 qmake -o Makefile v_rep.pro, 過程中將取得 Makefile.Release, 可利用 mingw32-make -f Makefile.Release 開始進行 V-rep 核心的編譯流程.

編譯結束後, 可以在 v_rep/release 目錄中得到 V-rep 系統執行所需的 v_rep.dll 與 libv_rep.dll.a.

開放源 Robot 相關模擬套件

V-rep: http://www.coppeliarobotics.com/



Webots: https://cyberbotics.com/



FreeCAD robot workbench: https://www.freecadweb.org/wiki/Robot_Workbench



Blender: https://www.blendernation.com/2018/12/11/robotic-design-with-blender/