Process

Once we decided to make a mechanical flower inspired by the arduino project by circuit.io (see acknowledgements section) , the design process began with testing different linkage lengths in Solidworks sketches to get the range of motion we wanted, and then printing a 2D slice of the petal linkage to test the motion, range differences compared to the model, tolerances between printed pieces, and tolerances with the steel pins. All the parts were designed from scratch, with an extra layer of larger petals added on the outside. A spring taken from a broken pen was used to have the mechanism function like a single-acting piston, as we knew we would want the flower to be triggered via a string pull from the base, to allow for a stem that was any length/shape. As we knew we would want the flower to light up at some point, natural PLA was chosen for our printing material, with infill reduced wherever possible to increase transparency and diffuse light.

Once enough adjustments were made for the 2D prototype to work well enough, full petals replaced the 2D test slices, and the entire mechanism was mounted on a white straw to act as the stem. A single Flora RGB Smart Neopixel was mounted within the flower, and the flower adapter to the straw was designed to allow the 3 wires from the neopixel to feed from outside into the straw. Unfortunately since the mechanism wasn't designed with housing wires in mind, required some glue, and took a lot of time to assemble, the center pieces weren't replaced with ones with holes, so the 3 wires are still exposed a bit on the outside of the flower.

With the flower was complete, we began making a breakout board for the argon as we knew a breadboard was a little too big for the footprint we wanted the base of the pot to have. The breakout board has a single row of female headers for all of the GPIO pins, as well as a power/ground row of headers, with the ground pins soldered on the back, along with the power and ground wires for the dc barrel plug adapter.

Afterwards, a platform was designed to mount the servo vertically so that the wire to actuate the flower could be centered in the pot while allowing the servo to fit within the footprint, and a neopixel ring holder was designed to mount on top of the servo to illuminate the walls of the flowerpot. We picked a decently large servo rather than using a microservo in case the friction between all the links of all 12 petals was too much for them to hold, and because we didn't want to make the servo work to burn out from working too hard as the closing state of the flower is the default, and requires constant actuation of the servo.

With the servo mounted, some spacers were used to space the top plate above the neopixel ring using 8 sets of long M3 bolts, with a lot of weight relief cut into it to reduce printing time. When attaching the flower to the straw and top plate, the fishing wire had a hard time passing through the straw, so nuts and bolts were used as weights to allow it to fall through the straw, before being removed after the neopixel wires were passed through as well. The wire was wrapped around two screws on the servo horn to allow the string to be pulled down when the servo was actuated, and hot glue was used to hold down the neopixel ring, some of the wires, and the dc barrel plug adapter to ensure that they wouldn't move much.

The last parts to be designed for the flower prototype were for the walls of the flower pot, as we wanted to ensure everything would fit first before it was all closed up. They were modeled after generic terra cotta flower pots, with the main lip at the top printed separately to remove the need for  support material and reduce printing time. It also served to reduce the height of the pot, and ensure that the rest of the assembly would have enough space to assemble properly as the barrel plug adapter sticks out and required the whole thing to be slid in at an angle to have enough clearance. The lip itself was further split into 3 pieces- an outer ring with an inner diameter larger than the rest of the pot/stem plate so that it could be slid on at the very end, and two halves of a ring that could be assembled from the sides that would be sandwiched by the stem plate and pot walls and provide mounting points for the ring to bolt into. These two halves had a lot of hexagonal weight relief cut into them as well to reduce printing time since they basically acted like large spacers.

For the sonar sensor part, we first built the basic circuit and got the proper range of distance that represented presence. Then we designed the logic of our focusing flower, it will detect the distance between laptop and user every 30 second, and if it keeps detecting presence for 5 minutes, the flower will open five degrees. In the testing process, we found that ultrasonic sensor sometimes will give some invalid/strange results, such as 98cm. To make our program more flexible, we adjusted the program, and prevented the counter from being interrupted by just a mis-measurement. 

We created two webhooks, one is used to pass servo value to the flower to open/close it based on focusing time. The other one is used to notify that button is pushed on one side, the LED on the other side should be lighted up.

Then we integrated the code on two sides, adjusted the range of servo, with proportional to focusing time.

Then we improved the circuit and made it much clearer and simpler, to enable it to fit into a wrapper box which can be nicely put on the laptop.