Since last week, our plans for the control tower have changed yet again. Following difficulties with reliably detecting ellipses from the hot pot and strainers, we decided it would be best to remove complexity from the already daunting computer vision algorithm and replace it onto the physical design aspect. The camera angle and position will again be fixed with respect to the hot pot and the touch screen will detach from the control tower in order to prevent accidental nudging of the camera while a user is poking the touch screen. This will prevent adding another huge development task to the computer vision algorithm on Christina’s workload, but it will cost Shane some work to figure out how encase the touch screen separately from the control tower and still elegantly run the wires. Right now, we are also thinking of adding a layered silicone placemat that will go underneath all of the components and the hot pot with trace-like slots for all the wires to run underneath in order to aid the user in setup and ensure correct placement of the motors and camera.
We have also potentially acquired a 3D printer that we can use instead of TechSpark to fabricate our parts. Christina will be bringing it to Pittsburgh on Monday, after which we can see if this kit-built printer is operational. Our plans right now are risky due to the large amount of 3D printing we want to do with the control tower and an additional detachable touchscreen enclosure. If the printer doesn’t work, we won’t be able to print at the cost of material and it would be too expensive to use TechSpark; we would have to scrap any 3D printing designs and find a new enclosure solution from scratch.