We’ve come a long way from the drawing board to arrive at our current prototype. Below is a detailed recount of how our design evolved, as well as the ideas, considerations and brainstorming processes which guided us along this M and T journey.
Stage 1 – Initial Design (and its problems)
Below is a diagram showcasing our initial design, with the head and length-sorting segments labelled accordingly.
For our first design, the head-sorting segment made use of the screw pushing mechanism which we had discovered in our “experiment”. The segment consisted of a slanted platform, an adjustable slider, and a pushing device as specified in the picture below. The angle between the slider and the platform will be adjusted such that it will only trap screws of the specified head size. Thereafter, the slider will move upwards and rotate these trapped screws into their correct orientation (head facing away). When the trapped screws reach the edge, their heads will rest nicely on the edge, while the screws that were not trapped will be fall over onto the other side of the platform and into an overflow box. The pushing mechanism to the left of the platform will them push the screws (with their heads still resting on the edge) horizontally such that they are transported to the length-sorting segment. This action will not be able to sort all the screws of a specific head size in one attempt, hence the overflow box will deposit the unsorted screws back onto the platform further sorting. Once all the screw for a particular head size are sorted, the angle of the slider will be adjusted accordingly to trap and sort screws of the next head size. We had initially also intended for the user to control the movement of the slider and the pushing mechanism via a phone application, which will be linked to an Arduino board via a Bluetooth connection.
Nonetheless, we soon discovered some major flaws with this design even before 3d printing:
– The trapping and rotating of screws into their correct orientation was much less consistent than what we anticipated. It does not work on many surfaces and requires a very specific application of force, something that was difficult for a stepper motor to achieve
– The large variations in the sizes of screws made the trapping and rotating process very complicated. Large and long screws (such as M6 80mm screws) would almost certainly get in the way of small and short screws (M2 4mm screws).
– It was extremely difficult to get small and short screws to hang with their heads resting on the edge of the platform, as the area available for hanging is very small. The interference from larger screws as mentioned earlier also meant that any small screws already hanging could be easily knocked off the platform.
– The process of the overflow box pouring the rest of the screws back onto the platform will have to be repeated many times, and users do not have an accurate way of determining when one head size of screws have been sorted and they can move on to the next head size
– The positioning of the pushing mechanism (to transport screws from the head to the length-sorting segment) and the overflow box are too close to each other. As both are moving parts that required stepper motors, this could lead to a clustering of wires making it very difficult to construct.
While we had hoped to stick to this design since the trapping and rotating of screws was our own “discovery”, we ultimately realized it was too inefficient not worth the effort to continuously modify it. Hence, we had to revamp the head sorting segment almost completely.
The main principle behind our length-sorting segment remained roughly the same, though it has undergone several modifications to function more reliably and consistently. Initially, the segment consisted of a single board tilted at an angle and sloped downwards. The sizes of the holes on the board are arranged in ascending order from left to right, and they correspond to the specific lengths of the screws we are sorting.
As the screws roll down the board with their heads resting on the edge, they will fall into the holes tip first only if the hole is long enough for the screw to fit in. For instance, the shortest screw will fall into the first hole from the left, but longer screws will continue rolling until they reach their corresponding hole. Containers would be placed beneath these holes to collect the falling screws. We tested this with a piece of cardboard in the video below. A ruler was attached to act as a smooth, consistent track for the screws to roll down.
There were two main problems with this length-sorting design:
– The speed of the screws rolling down the track was too fast and erratic. Many screws simply fell off the track before they could reach their allocated hole. (Video below)
– There were a few types of screws whose head size diameter was greater than its length. As we initially cut the holes in the board to correspond exactly to the length of the screws, the holes were not big enough for some screw heads to pass through. This caused a blockage in the holes which disrupted the sorting process. (Video below)
Stage 2 – Brainstorming for a new head-sorting segment
Stage 3 – Refining the length sorting segment
Stage 4 – Connecting the segments
Stage 5 – Wiring
Stage 6 – Coding
Stage 7 – Assembling the entire structure