STEP 1: WE NEED A CONCEPT
Building a little Mars rover may be a high- flying idea – but the real implementation needs a realistic and affordable concept.
How big should our rover grow? Building a very small rover makes it harder to integrate all features, building too large makes it more expensive, difficult to handle and may require stronger materials for body and undercarriage and expensive drive motors. On the other hand a bigger rover has better off-road capabilities.
My rover will have a body size of 36 x 50 cm, the overall length will be around 70 cm and the rover will be half a meter wide.
The undercarriage will be a simplified rocker-bogie system with a 6x6 drive. The wheels are not steerable, we will turn the rover by running the motors forward on the one side and backward on the other side. The disadvantage is that the steering will be not very precise and driving to a target point will be more difficult. The advantage is the simple and cost- effective design. Since we mount the drive motors directly to the wheels, we need gearmotors to get the RPM down.
The robotic arm
There will be very simple robotic arm - mainly because the original rover has a manipulator arm which carries many scientific instruments and cameras.
The rover will have a camera mounted ontop of a mast to be able to oversee it’s surroundings. Furthermore we will use ultrasonic sensors to detect obstacles very close to rover. An IMU may be used to help to guide the rover, but I’m not yet sure about this.
Let’s keep it simple: We will use batteries.
My concept is to use two computers: One to control the motors and servos to drive the rover and to move the robotic arm and the camera mast. The second computer is the main computing element and will be the brain of the rover: It needs to read data from the camera and the ultrasonic sensors to build a map of the area the rover is facing on its way to the designated target. The computer will then decide autonomously which route the rover will take to get to the target.
Operating the rover
We will not remote control the rover! Due to the long turnaround times for a signal from Mars to earth and back it is impossible to joystick a rover on the surface of Mars. In reality it works like this: An operator on earth sends the rover a command like: “Drive to a target point 200 meters ahead and 50 meters left from your current position”. The rover’s main computer receives this command and starts to plan a bunch of operations how to get there. E.g.: “Drive 5 meters forward, turn 45 degrees left, drive another 5 meters, turn right and so on”.
Then it will send directives to the secondary computer to get the operations done. These commands would look like: “Start all 6 motors running forward with 50 rpm. Stop after 10 seconds”.
To be able to send commands to the rover and to receive data from the rover, we will use a "basestation” software. This software will be coded in Java and is therefore runnable on any Linux / Windows / Apple computer or tablet.