Choosing computational device for a guided vehicle

Community General Discussions
1

I'm building a robot that in the hope of creating an automated guided vehicle. Does anyone have any experience on the choice of computational power for such a robot? The setup I was thinking about using was to base it around a Raspberry Pi. The vehicle will have to interact with a computer wirelessly and also allow itself to be controlled from the vehicle. Also I hope to utilise a known map for navigation purposes.

Is anyone able to point me in the right direction for reading material to learn about this or explain the benefits of different computational methods?

Any help is very much appreciated.

2

A history lesson –
The Apollo Guidance Computer (AGC) was fairly basic by today’s standard. It had approximately 64Kbyte of memory and operated at 0.043MHz. And yet, this ingenious basic computer was able to ferry astronauts from the earth to the moon and return them safely. Read more: http://en.wikipedia.org/wiki/Apollo_Guidance_Computer#Applications_outside_Apollo

System design specification:

  • Laptop (ground station:uplink/downlink comms, rover telemetry, etc)
  • Raspberry PI model B (rover mobility, navigation, etc)
  • GPS module
  • 2 Zigbee Pro RF modules

    • To jump start your project more information can be found here: https://www.robotshop.com/letsmakerobots/node/18922

3

Further Detail

Many thanks for all the swift responses. Apologies that my first attempt was a little bit vague. I didn’t want to write too much as this for a university project so I’m trying to learn about it all without just asking for all the answers as I’m really keen to develop an understanding of this field.

The AGV is to carry a small object around a warehouse/factory floor environment. A maximum mass of 10kg will be transported. The operatable area has marked pedestrian walkways which the AGV will use as its path to move around. However, in the event of a static or dynamic blockage/obstruction (such as a pedestrian), where possible I want to allow the AGV to navigate around the object, or if not possible, then to follow a different path. This was the reasoning behind using a raspberry pi as it would allow for that computation to be done from the AGV and also provides a system for which software can be written.

The AGV should be able to be given a destination by an operator either remotely or from the machine, and demonstrate an ability to be wirelessly controlled by a user. Thus the wireless capability is for this reason.

What would you initial thoughts be here? Is a raspberry pi a good route to take? And if so, is it fairly easy to link it via  bread board to control a motor?

I really appreciate your help, so many thanks for it all.