An important choice when designing your own Rover-Pi robot is to decide if you want to make it fully self-contained, or if you are happy to have a tether (a long control/power cable connected to the Rover-Pi). Using a tether, you can keep the weight of the Rover-Pi robot down, which means the small motors will be able to move the unit with ease. This will allow you to keep the Raspberry Pi separate from the main unit so that it can remain connected to a screen and keyboard for easy programming and debugging. The main disadvantage is that you will need a long, umbilical-like connection to your Rover-Pi robot (with a wire for each control signal) that may impede its movement. However, as we will see later, you may only need three or four wires to provide all of the control you need (see the Using I/O expanders section in the next recipe).

If you intend to mount the Raspberry Pi directly on the Rover-Pi robot, you will need a suitable power supply, such as a phone charger battery pack. If the battery pack has two USB ports, then you may be able to use it as a power source to drive both the Raspberry Pi and the motors. The unit must be able to maintain the supplies independently, as any power spike caused by driving the motors could reset the Raspberry Pi.

Remember that if the Raspberry Pi is now attached to the robot, you will need a means to control it. This could be a USB Wi-Fi dongle that allows a remote connection via SSH (and so on), or a wireless keyboard (that uses RF/Bluetooth), or even the GPIO D-Pad from Chapter 9, Using Python to Drive Hardware, which can be used for direct control.

However, the more you mount on the chassis, the harder the motors will need to work to move. You may find that stronger motors are required, rather than the little ones used here. A Rover-Pi robot powered by a USB battery pack is shown in the following photo: