Siduri – An Arduino Control Smart Coaster


Siduri named after the Sumerian god of happiness and merriment, is a smart coaster for your drinks. It recognizes when a glass is near empty and then glows yellow to alert waiters that you will be in need of a refill soon.

Designed specifically for lounge and club settings, Siduri helps nightlife revelers politely draw the attention of barmaids and helps bartenders to keep the drinks flowing.

The coaster is powered by Ardunio and uses an FSR sensor that recognizes the difference in the weight of a drinking glass. A button located near the base of the coaster allows bar staff to calibrate the coaster to an unlimited amount of drinking glass types into the coaster’s memory. Hidden under the white acrylic top are three surface mount LED’s that breath a soft yellow light when the FSR sensor recognizes when a glass is 3/4 empty. The remaining materials were laser cut out of 1/8” wood to give the coaster a manly, yet classy aged feel.

If you would like to make your own smart coaster, here is how I did it…

Link to my youtube video:

WaiterLite Smart Coaster Update


Since we last met I did a lot of field research. I visited several bars. I shared my prototype with bartenders, got some feedback and made many observations about my audience. All of which improved on my product design (see the above).

The physical prototype is still a work in progress but I learned alot about how to improve my design. Below are my learnings so far:

The Bartenders

2013-11-17 23.49.01

This is Rebecca. Rebecca is a bartender at Bedford Hall.  She told me that as a bartender she can’t attend to everyone, so it’s more important that the wait staff that takes drinks over to the tables be her eyes and ears.  She thinks the smart coaster idea is more important to the waitresses than to her because its their job to go over to the people and bring drink orders back to the bartender. Rebecca also suggested that I move the LEDs from the outside of the coaster to the middle of the coaster.  Her feedback was that the LEDs could be distracting in a dimly lit space, so its better to defuse the light through the drinking glass.

The Glasses

2013-11-16 22.39.18IMG_0107

After visiting several different bars I learned that I have to account for six different kinds of drinks glasses. There is water, wine, beer, martini, whiskey and a mixed drink glasses. When placed on my FSR sensor each register a different weight.  This presents a challenge because in my current arduino sketch the bartender would have to calibrate the coaster to each glass.

My current design solution is to calibrate the coaster to recognize all the weights. I did this by storyboarding the optimal use case scenario of my coaster.


IMG_0103 Puesdo Code

I learned quickly that it is far easier to show how to use the coaster than to explain it.  So I sketched out what the ideal scenario would look like when the smart coaster is in use.  I also found this technique helpful when it came to re-writing my code.

When planning out how to revise my code I sketched out what each step should look like. This is slightly different than sketching out a storyboard for users. Sketching out code is called Puedo Code.  It maps out each technical step of my use case scenario. For example “flash yellow LED’s when the coaster is processing”,  this helps me understand how I need to write code to meet these conditions.


Coaster Components 2013-12-02 12.43.03

After taking in lots of feedback, I began to plan out how I wanted to revise my concept. the above photo shows how I began thinking about revising my cardboard prototype into wood. I used a laser cutter to cut out my materials.


Currently I am considering different way to reduce the size of my concept.  As is the form factor is to large.  I found some sample projects on instructable here and here I want to take a look at.  Revisions coming soon.


Waiterlite Coaster Proposal

For my final project I want to revise my waiterlite coaster.

The initial concept was designed to alert wait staff when patrons need help getting a bartender’s attention during high traffic times such as happy hour. I would like to revise and build on this concept.

The challenge I face with my current sketch and prototype is that my FSR sensor only works for specific type of glasses. This is because my current sketch is design to recognize the weight a single type of cup. Each time a new cup is placed on the sensor I have to reset the sensor to recognize new thresholds for each new glass.

My next steps will include calibrating an FSR sensor to automatically recognize when a glass is full or empty.  I will attempt to revise my arduino code using motor limits to automatically recognize the upper and lower thresholds that a glass will exert of the coaster.

I also would like learn how to use the CNC machine so that I can cut out a nice coaster made out of wood.  Ideally the coaster will be designed to house and conceal the flora board and the battery pack in single form factor.

Next Steps:

Target Audience

Although the concept is designed to help bartenders to quickly recognize when users need a new drink, they are not the true target audience.  The real challenge is to identify a bar that is not able to meet consumer demand during volume.  Three places I am considering are places with

Sweet Science



Sketch Concepts

Explore ways to create prototype. Buy materials like cork and wood and see which materials are work best.

The Waiter Coaster

A coaster with LED light indictors that notifies waiters when you need service.

IMG_0179 IMG_0177

I used an FRS censor to create three thresholds that identify when a class in full, half drank or empty. Within the arduino code I used several “if else” statements that distinguish between the weight of the glass.

IMG_0174 IMG_0182Then I used a glue gun to and wooden sticks to connect the two parts.


The FRS censors are calibrated to recognize three thresholds or weights.  When the glass is full a red LED indicator turns red. When the glass is half full the LED indicator turns tell. This tells a waiter that you are ok and don’t need help.  The final LED state turns green when the glass is near empty, indicating that you need a refill.

Button State Exploration

This assignment was fun. Through exploration, trial/error I learned that the Digitalpullup button needs a resistor in order to recognize the full circuit.  When a resistor is not used the flow of energy circles around in a “high” state from the LED to GND. A resistor is need to resist the flow of energy to return to GRN enabling the LED to light up.

A seconding finding for me was within the StateChangeDetection sketch.  Defining input and output is crucial to defining the state of the LED (on vs. off).  The pushbuttoncounter is also important and dictates how many “clicks” it take to turn the LED on and off.  I also found that if you added-in a delay feature you can change the timing of when the LED turns on and off.



Automatic Fuzzy Pillow Light

Automatic Light Pillow

A travel size pillow that contain two LED lights for reading.

I observed that most people that own a travel pillow like to read. Often these people are in transit and don’t have good reading light. This pillow is intended for those that like to read in the dark.


2013-10-08 12.17.482013-10-07 11.06.17

I began the project by creating two sketches.  The first is a concept sketch where I tinkered with a form factor design and I figured out what electronic components I would need to bring this idea to life.  The second sketch was why electronic component sketch.  I outlined how I wanted the circuits to flow and designed a template for how I would fit the cables into the housing.

Wiring and Testing

Next I bought several types of LED light, a breadboard to test proof of concept2013-10-07 12.37.58 2013-10-07 13.29.05


2013-10-08 12.03.45 2013-10-08 12.03.56

Then I went to Mood Designer Fabrics in the Garment District with an actual travel pillow.  I matched the fleece and spandex material and bought 1 yard of each. Then I created an outline of the pillow, cut out the material and then sewed them together.  Finally I stuffed the pillow and added all of the electronic components.

Next Steps

Given more time, I would like to use a laser cutter to cut out the material precisely and them take the materials to a seamstress to sew the seams together properly.  I would also like to create an enclosed housing fixture for my flora board and battery pack so that I can swap out the components as need be.

I’d also like to explore using proximity sensors that automatically recognize gesture behaviors like the opening of a book or the turning of pages as a trigger for the LED lights