Universal Vision Sensor Mount
Entry ID #: 6619
Created: Tue, Jan 15, 2019 12:14 PM
This year, VEX introduced the new V5 control system for the High School Competition. Since the beginning of this season, we have had a great time beta testing the new V5 control system, learning about it’s quirks and abilities every day. One aspect of the new system we were really looking forward to was the vision sensor. Over the summer, we heard numerous rumors about this “camera” and how it could recognize colors. In our heads, we saw endless applications of this new sensor to the VEX ecosystem. Right off the bat, we found the vision sensors extremely useful— especially in this year’s Turning Point game. With each flag being 6 inches by 7.2 inches, we realized that it would be advantageous to have a robot-flag auto alignment system using color detection on the vision sensor. As we eased into testing the vision sensor, we ran into an issue in figuring out the best angle and position for it. Often times, we had the camera plugged into our laptop while angling the sensor; at certain angles, the sensor had trouble detecting the color combination of the flags. In addition, we would often use our hands to hold the vision sensors temporarily, causing inconsistencies. After finding the optimal angle for the vision sensor so that the sensor had coverage of the flags, we decided to bend a VEX plate to the exact angle required. After spending time mounting the vision sensor on the robot, we continued testing the alignment and realized that the angle we had just created was not desirable because the angle would mean our robot would have to drive quite close to the flags in order to detect the flags’ positions. We wanted to be a bit farther away so our robot’s shooting range would not be limited to the range of our vision sensor. This change meant bending new plating. This time consuming process inspired us into thinking of a solution to easily mount the vision sensor at any angle we wanted! The mount we prototyped was inspired by 2 existing VEX parts: the shaft collar and the hinge. We wanted to make the mount capable of angling up and down as well as rotating left and right to adapt to any VEX game. The base plate of the mount has 4 holes (one per corner) that can be easily mounted to VEX channels with regular 8-32 screws. The circular core of the base plate mates with a circular mount that fastens onto the base plate. We wanted to find a way to maintain rotational position at any angle. This is when the shaft collar came to mind; the shaft collar uses a set screw to hold it’s position along an axle. We thought that in this case, the shaft collar would be able to serve the same purpose. For optimal stability, we made the ring larger so that the sensor would not wobble when fastened. The circular mount fits nicely onto the circular core of the base plate, rotating freely. On the side of the circular mount, we put 1 screw so that once the position is set, the screw is tightened, biting into the center core of the base plate as a shaft collar does. Next, we looked at the VEX hinge and created our own version. Unlike the VEX hinge, the hinge in this case would need a way to maintain any position we desired. We decided to use parts already available to us for the hinge’s pivot: 8-32 screws and standoffs! When fastened, the hinge would maintain its position. This mount would make it simple to set certain positions of the vision sensor. With a changeable angle, it would save time and parts; instead of having to bend new metal each time we wanted a new angle for the sensor, our universal angler would allow us to change the angle quickly anytime we want. Through our experiences at numerous regional competitions, we believe that this issue is common across all V5 teams integrating the vision sensor into their game. Teams have come up with elaborate ways of mounting their vision sensor, requiring immense customization. The mount we created would eliminate this issue, as it is a compact mount which can vary angular and rotational position. We wouldn’t have been able to implement our prototype without CAD software. Autodesk Inventor allowed us to design our idea on the computer. After 2 iterations of 3D printing, we were able to settle on the final design you see attached. Ultimately, this challenge taught us the importance of utilizing the design process in every project we pursue. First, Angela on our team came up with the idea. As a team, we continued to refine her idea and develop it through sketches. After developing the idea, we analyzed the dimensions of the vision sensor to understand what our parameters were. To computerize our idea, we used Autodesk Inventor to design and visualize the mount (to scale). Visualizing the mount allowed us to prove the concept virtually. We then continued, 3D printing the mount and physically implementing it on our robot. However, our first design did not work as planned because we made the holes for the sensor too close to the pivot of the hinge, making the vision sensor unable to fit on fully. Through this, we learned to really stress doing “background research” in understanding the tolerances and specifications of our design. This made us realize the true importance of the design process and its effectiveness in projects. On a competition robotics team, CAD software allows teams to plan out their robot entirely at first before building it. Doing so would allow teams to eliminate as many foreseeable problems as possible. Learning CAD as a team will help all members as future engineers with the growing popularity of rapid prototyping.