A common problem home bakers often encounter is having either too much or too little flour due to it being compressible. The Smart Dispenser is a tool designed to measure out particulate materials for baking such as flour and sugar by mass, eliminating any uncertainty in measurements. The Smart Dispenser was a 7 week project with rapid prototyping and multiple iterations.
The first prototype filled the top with flour. A potentiometer was used to control a servo to open and close the funnel opening. The first prototype allowed for some control dispensing the flour, but it there was an issue with the amount dispensed and quickly being able to seal the hole. Finally, with a larger amount of flour, it gets packed at the bottom of the funnel and is unable to be dispensed.
The load scale gave a live reading about the mass of material on the plate, allowing for the hatch to close itself when it reached the target weight.
Following the lessons in week 1, the entire system was redesigned. To avoid the flour compressing and becoming a "solid," the funnel was replaced by a silo. A rotating disc with a plate pinned to it was placed under it. As the disc rotated, the plate would be pulled back and forth under the silo which released material and then sealed it. For this week, the parts were created in SolidWorks but physically created with cardboard and materials on hand. Since the silo opening is controlled by half rotations of the wheel, there is a minimum amount of material dispensed assuming a constant wheel speed. This problem was addressed in prototype 4.
SolidWorks CAD of wheel, pin, and plate in the open and closed positions.
This prototype was made out of the same popsicle sticks, wooden dowels, DC motor, and cardboard as the jumper robot.
To avoid having a large minimum amount of material dispensed, slits were added to both slides of the plate. Limit switches were added under the plate, and when both of them were under the slits, it represented the silo being closed. When both limit switches were not actuated and the final material weight was satisfied, the system was stopped. A paper clip was also added on top of the limit switches to prevent the plate from being caught in the slits. When testing the week 4 plate, material got under the rails, causing extra friction. In week 5, the gaps were narrowed and the system was nearly fully assembled.
Top left: limit switch with adjusted actuator
Bottom left: Plate iterations
Video: Assembled system without load scale. The plate was programmed to stop on a timer in this video demonstration.
Unfortunately, videos of the final prototype were not saved because it was an in-person demonstration at the presentation fair. Prototype 6 changed the plate to only have one slit and one limit switch. The limit switch was placed at the far end of the plate for the sealed position. When the mass on the load scale was attained and the limit switch was hit, the system would stop. This was a much simpler and elegant solution to having two limit switches that frequently jammed and had enough resistance to bow the acrylic plate. Finally, as the target weight was approached, the motor driver sped up the wheel to give a more refined amount of material.
Stepping back from the project, I recognize some fatal flaws with the final design of the project. First and foremost is the seal between the plate and the silo. Without a perfect seal between the silo and the plate, material will always go underneath and sit on the plate which increases friction with the plate and silo. On the other hand, with a perfect seal, there is too much resistance for the motor to move the plate. The Smart Dispenser was also very messy. Cutting the flour and sugar with a moving plate back and forth not only shot it out horizontally under the plate, it also launched the material over the silo!
For the next steps, I would redesign the project again. Instead of using a motor to drive a plate, I would use a motor to drive an Archimedes screw and use that to drive the motion of flour. They are currently being used in feed augers for agriculture. In a smaller scale and after multiple prototypes, it may be an effective alternative.