Open the Call for Multipilot 32 Developer

We NEED C / C++ DEVELOPERS and Beta Tester.

if you are interested to the project contact me pm and add your comment in this thread.

Multipilot32 will be first processor of series that will Support VROS.

VROS will be Operating system choose and developed from Virtual Robotix community for  the next generation DIY and Professional Robotix drone.

VROS will be as ROS is for High End processor and could be fully interface  with ROS for improve his advanced functionality as frontend towars low level electronic interface.

It will be :

  • Simple.
    • using C++ as main language.
    • Implementing a simple api language like a Wire or Basic.
    • Have a lot of example available of ready to use code.
    • have a powerfull and stable library fully documented.
    • have a large and international community of users.
    • Supporting Scripting language.
  • Realtime.
    • Use a Powerfull Operating system that support.
      • Preemptive multitasting.
      • multicore.
      • predictive
      • HAL : Hardware Abstraction Layer.
        • I/O Port interaction
        • USART
        • SPI
        • I2C
        • PWM
        • USB
        • PAL or VGA video out (low res for osd application).
        • SD File Systems.
  • Supporting Scripting language
    • Use scripting language as Basic for fast debug and assembling advanced api function or precompiled task.
  • Standar API for Robotix application , fully documented with a lot of  example.
    • Radio reciver control.
      • Standard PPM Input
      •  Multi channel
      •  PPM SUM 
    • Servo Control.
      • Standard and Hires PWM Output (16 Bit PWM).
      • I2C and CAN bus servo Protocol.
    • Brushless Controll
    •  
      • Standard and Hires PWM Output (16 Bit PWM).
      • I2C and CAN bus servo Protocol.
    • AHRS Algorithm and DataFusion.
      • PID
      • DCM
    • HAL for a lot of different sensor.
      • the main idea is that the api interface is indipendent from technology sensor. So if you use an i2c gyro or analog gyro you change only driver not the high level application.
    • Telemetry standard protocol.
      • Based of RPC approach , binary with crc and flow control , optimized for low band transmission. FSK 300 baud too ...
    • Scripting language available for differente task :
      • Configuration.
      • Logging production.
      • Fly control mission.
  • Portable on different hardware platform.
    • The VROS will be support :
      • 8 Bit Processor as AVR or PIC
      • 32 Bir Processor as ARM .
  • Available opensource and professional support and license.
    • That is good for some application where some application start as experimental low cost prototype and became professional application after a while.
This is my first brainstorming ... I'm waiting your suggest and proposal.

You need to be a member of FOXTEAM UAV CLAN to add comments!

Join FOXTEAM UAV CLAN

Email me when people reply –

Replies

  • Hello everybody,

    Roberto invited me to post on this blog and present myself. Just few words about ChibiOS/RT (my project), it is a small real time OS and should not be compared with Linux or other general purpose operating systems. It is best used in simple tasks where the response time is an important parameter, for example:
    - Motors control.
    - Balance.
    - I/O handling.
    - Sensors handling.
    On an STM32 you can realistically handle events in the range of few tens of microseconds.

    Higher level tasks like navigation, obstacles avoidance, image/video processing should be handled (as others suggested in this thread), by more complex, higher level, processes running on a more general operating system. My idea of architecture would be composed, by one or more, relatively slow, high level processes running under Linux communicating, using some kind of bus/protocol, with one or more peripheral processors running a RTOS like ChibiOS/RT handling the critical and simple parts.

    About the central processor, I was Atom boards were discussed by just this morning I found something that might be interesting:

    http://www.igloocommunity.org/

    I really like this board, it is powerful and packed with sensors, basically it is the hardware of a last generation smart phone, it could be a worthy alternative to the Atom, probably it would draw less current too.

    Giovanni
    • Giovanni said:

      > I found something that might be interesting: http://www.igloocommunity.org/

      See also the Openmoko mobile phone project (Linux based), they are currently working on the new GTA04 hardware (project page here).

      Sorry Roberto, but yes, smartphones in general seem to bring much that is required for UAVs like sensors, microcontroller, and even GPS, WLAN, Bluetooth, camera, plus a display and Internet access (GPRS/UMTS). The latter would even let you to stream live video to the Internet...

      Not sure though how to connect ESCs/servos to such a beast. So a dedicated platform like Multipilot32 seems to be a more practical approach.

      BTW, the AR.Drone does it the other way around: Here an iPhone or iPod touch is used as a remote control (watch the video on the product homepage).

      Stephan

    • FYI: Please see also http://www.virtualrobotix.com/forum/topics/multipilot32-need-a-rt-o...

      Stephan

  • Road map for future development will be interesting, thank you.
  • Hi Michael,

    Eurotech is an Italian company based in part of italy called friuli. So I think they produce Atom board in u.s.a. I have a good relation with top managment of Eurotech and they told me that are interest to support our project. When I have news and rodmap for future development if you are interest i inform you about the evolution of project.

    Regards

    Roberto

  • Not sure if I fully understand your approach. Assuming by ROS you mean www.ros.org then you still need some underlying operating system on which ROS can run. Correct me if I am wrong but AFAIK, ROS currently runs on unix-like systems.

    There are quite a few embedded (real-time) operating systems out there, which I would prefer, e.g. VxWorks, QNX Neutrino, or the mini OS FreeRTOS just to name some. However, the first two are commercial and require a license and they are more like hydra-headed monsters whereas FreeRTOS is just a very basic OS.

    Maybe I get things wrong here but what would be your starting point for VROS? Writing an RTOS from scratch is not an easy task, believe me.

    • Hi Stephan,

      I'm testing ChibiOS/RT the link is : www.ChibiOS.org.

      The developer of this RTOS start is project in 1992 and he are interest to work togheter with us at VROS project.

      http://www.virtualrobotix.com/profile/GiovanniDiSirio?xg_source=pro...

      Our Idea is to develop a Distribution that include : Ide , ChibiOS/RT , VR API.

      My idea is that we develop a connector with ROS application framework. Actualy we're working on ROS Slam .

      Regards

      Roberto

This reply was deleted.