Robot Design Proposal
In developing our robot, we’ve looked at what would be beneficial in a real life situation. We’ve considered what would be necessary in real-life lunar exploration, and have tried to design a robot which would simulate capabilities a lunar rover would/ could have. One of the main aspects we feel would be needed to reach is overall control. With control comes being able to see everything and include features which assist the controller (help them to complete their mission). Say an object on the moon has already been discovered, but the robot operator doesn’t know what it is or doesn’t even recognize it as a “find”! Making some elements of the robot and software operate autonomously would help lessen operator error while still allowing the operator to move along in the mission. This way, they could focus on the ultimate goals of the mission and still maybe make a few additional discoveries. A possible solution would be to have a categorization system that would identify certain objects that have been recorded before. This system could use an image recognition device or something of the like. It would then ‘know’ to take a photograph of the object and organize it accordingly within in the data collected. We plan to simulate these abilities with the use of sensors and camera mounts which permit full field coverage.
For this to be accomplished, we will recreate a panoramic view of sorts. With the use of only one camera, we will create a unique arm which can rotate the camera a full 360 degrees as well as adjust the angle. This provides so much more visual range (which, again, would provide so much more to a lunar expedition), and it’s as simple as a two motor mount! The camera would be mounted perpendicular to one motor (rotating the base clockwise or counter-clockwise) and the other parallel (rotating the camera up and down). We also plan to have the whole system mounted on a stand which could extend upward or lower itself, adjusted for the very best view. In real-life, a system such as this would be controlled manually, but we feel it could also assist in our autonomous proposal. The image recognition system would be activated and automatically move the camera to the appropriate position to take a picture. In our case, when the chosen sensors are activated, the camera and its mounts will make their move (may even use a color sensor depending on the aesthetics of the game artifacts). We also feel that, while a live feed of camera footage going back to a control station would be very important, storing footage on an internal memory card would keep the mission safe and recorded just incase network problems were to occur. So, we want to make sure the video is recorded internally, as well.
So for our mechanical design, we are going for a stable tank drive and tread system with separate, but connecting, joints (pivots). This will make our robot a bit more robust/ durable, and just help with overall stability. Even so, for extra security, we will create a self-righting system just in case the robot was to “get into a sticky situation”. In addition to all this, we feel our robot could become even more mobile if the central mass is able to extend upward. The top of the robot could rise up and go over objects rather than going around. Then, for full barriers, it could “stay down” for a low center of gravity. We want our robot to remain balanced so that the mission will not be compromised! We realize that it’s still good to have a back up plan, so another sensor (such as another bumper switch) could be positioned so that it is activated when the robot would topple [hits the ground first]. Another drive system would be activated to ‘right’ itself, and all is well!
Then, initially, we plan to use bumper sensors and possibly the ultrasonic sensors to help with navigation- as similar sensors can be used to help avoid obstacles or do what’s necessary to overcome rough terrain. Knowing the field layout, the sensors can be positioned in the most efficient structure. So even as a real lunar rover may track stars and the position of the sun for partial navigation- we are looking at smaller scale navigation for things a bit closer, especially for the purposes of the game.
