Author: aakinci

Personal Accomplishments

• Attached a sturdy 1 foot tall base below the established system in order to support the circuitry going beneath, elevated surface above the cup to optimize flow rate, and be aesthetically more pleasing
  • Here is a photo of it in the woodshop as it is glueing. Note that the weights are only temporarily there in order to make the glue as strong as possible.
• Tested the established system with various different requests including the products page and suggestion page in order to converge to our established requirements.
  • The valve selection was 100% correct using all of the products page drinks and requesting various suggestions as well. Tested 50 drinks and the correct valves would open each time.

Next Week

• This is our final week so its going to be presenting to the class and writing up our report. We have to do both of the report and the final poster to get it ready by next week’s demo

Schedule: On time

Personal Accomplishments

Scaled out the circuitry to support 6 valves and organized circuitry to make setup easier

Helped organize the tubes and containers by taping the valves to the wood and labeling/mapping ingredients based on our website.

Experimented with different valve pressures and measured amount of liquid dispensed to determine ball park estimate of ratio of timing to amount in ML

Below is the image of attaching the valves to the wood: https://drive.google.com/file/d/1NUxk5qAmYsefShDEVbOFaYRmuMOm26_l/view?usp=sharing

Below is the image of scaling out the circuitry:  https://drive.google.com/file/d/1aPcGPDX6_HwjfodIvGQHIX_9pdHxlQD1/view?usp=sharing

Verification

Since I was the lead contributor for the valve dispensing system, my verification and testing was in the form of ensuring certain valves opened according to ingredients chosen. Hence, my testing was experimenting with different combinations of drinks and timing how long the corresponding valves stayed open to ensure the circuitry was correct. Additionally, I testing different valve pressures and found that the pressure of 0.15 Pa was the optimal pressure for our system, I plan on doing more verification to ensure it fits in to the entire system.

Next Week Deliverables

Further testing of the dispensing system and overall system

Build the legs that will hold up the dispensing system and drill holes for the tubes and wiring the go through

Organize the entire system to prep for demo day

Status: On schedule

This week we began optimizing our system to look more visually appealing for the eventual final demo, now that we have a working system end to end. We began by shortening the pipes and taping the valves to our wooden structure so that the flow rate is consistent and the dispensing system looks cleaner. We also scaled out to 6 bottles and numbered them in certain positions on the dispensing unit that map directly with the website ingredients. Additionally, we improved the AI recommendation service to fully output customize drinks based on variety of user input and added in a “fun fact” which is additional information on the drink being made. Furthermore, we printed out the platform that the cup will sit on and the stirring mechanism will rest beneath. Finally, we scaled out the circuitry that attaches to each valve such that now we can control 6 different solenoids related to the 6 ingredients.

 

Below is the image of attaching the valves to the wood: https://drive.google.com/file/d/1NUxk5qAmYsefShDEVbOFaYRmuMOm26_l/view?usp=sharing

Below is the image of scaling out the circuitry:  https://drive.google.com/file/d/1aPcGPDX6_HwjfodIvGQHIX_9pdHxlQD1/view?usp=sharing

Below is the full image of all valves attached to the system: https://drive.google.com/file/d/1G9PB19FqwNt-jJ9HlFJwAHforfFz5lIG/view?usp=sharing

 

Verification and Validation

In terms of testing the dispensing system, we are currently in the process of testing amount of mL dispensed and mapping it to timings the valves are open to ensure the right amount of each ingredient is poured. We are testing this by individually measuring dispensing amounts  using a measuring cylinder and currently experimented with two different valve pressures to determine which one gives a steady and reliable flow. We also are planning to run several trials to observe and minimize leakage, using specialized safety clamps to observe where leakage is happening and avoid it. Finally, we are planning to do many different combinations of ingredients using the inbuild ingredients page and AI recommendation service to ensure certain valves correctly open. We plan on measuring the final amount of product and ensure it matches the sum of the ingredients also using the measuring cylinder.

In terms of the mixing system, it is pretty simple and reliable in its mechanism, so testing will be a series of repetitions in the overall system, ensuring that the pill spins for a certain number of revolutions for sufficient mixing. This number will be determined later after several trials of experimentation.

 

In terms of the garnish system with the robot arm, we currently have it fully integrated with the rest of the system but plan on testing individually to ensure the robot arm can reliably place the garnish in the cup. To do so, we plan on transitioning to the actual garnishes we are going to use for the demo, and test the robot arm on all the different ingredients multiple times, to ensure the robot arm and clamp is configured to pick up these unique shapes.

As a system altogether, we plan on running several tests using different combinations of ingredients, amounts, garnish selections to ensure the robustness of our system. We also plan on experimenting with different orientations to validate the quality of the system in an external enviroment.

 

Status: On schedule

 

 

 

 

 

Personal Accomplishments

• Implemented the products website that shows the curated drinks given the users recipes and images of the associated drinks. It is connected with the hardware so it is able to make the drinks directly
  • Git Commit:
  • Video: 
• Wrote the code for the AI speaker to say a fun fact about the drink and to repeat the user’s name. This will ultimately lead to a more personable experience for the user
  • Video
• Bought the wood 4×4 pieces that will be able to elevate the system above so that it can 1.) hide the circuitry from the user 2.) be above the cup so the liquid can flow downwards. I was going to build it this week, however, the wood shop is closed until the end of the next week. I’m hoping it will be open by the demo time so I can properly build it then.

 

Next Week Delivarables

• Have the table support fully built to support the system
• An accurate number for the system based on our metrics we established in the beginning

Status: On Schedule

This week we started to shift our focus from one simple drink working to a scaled out system that serves multiple ingredients. With this, managing the more valves and components introduces high chances of leakage, timing mismatches and hardware failures. To mitigate this risk, we created a valve system that been scaled to handle multiple valves with additional screws to minimize leakage. We also tested the integration between the wheel and arm using serial handshakes. A contingency plan will be reverting to fewer valves for the final demo if necessary.

Another risk that we had was timing and synchronization issues. The pouring of multiple ingredients in parallel introduces incorrect timing that can impact the overall latency of the product. To mitigate this, we incorporated multi-threading (concepts that we learned from Carnegie Mellon system classes like 15-440 and 15-418). Specifically, the threading was implemented to handle concurrent valve control with specified delays. Here is the video of that working

Another risk was the physical support and structural stability. Specifically, we had to create a wooden bookshelf structure was built to mount and support the valves. We will be adding 3D-printed components like the stir cup holder both for functionality and secure fitting.

The dispensing system was redesigned to support parallel valve control and expanded from handling 1 drink to a scalable system for 6 drinks. We did this because we need to support multiple drink combinations on demand.

Some other accomplishments we have is a fully working AI Drink recommender with voice integration. We also had issues with the robot malfunctioning. We were able to get the root cause of the error being a power and motor issue and debugged accordingly.

Currently we are still on schedule with the following tasks planned for next week

• scale up to 6 working drinks
• Integrate stirring and cup holder as base.
• Measure and calibrate flow rates for consistent pouring.
• Perform full integration testing
• Finalize hardware layout.

The following are deliverables from this week.

Parallel Valve Control + Integration Demo
Parallel pouring system now works using Python threading:
Code: Valve Threading Code
Video: Parallel Pouring Demo

Valve System Physical Expansion

Expanded system to handle 2 valves with plans to scale to 6.
Photo: Valve Setup

Wheel and Arm Integration
Components now communicate via serial handshakes.
Video: Integration Video

AI Drink Recommender with Voice Interaction
Voice-assisted AI bartender implemented.
Demo: AI Bartender Demo

Core Stirrer Functionality

3D printed magnet holders integrated with motor.

Demo: Stirrer Core Demo

 

Personal Accomplishments

• Implemented the AI drink reccommender for the website, including a voice that communicates with the user that simulates an actual bartender encounter:
  • Demo: https://drive.google.com/file/d/1p8YH_9dif7g0S69WDj-xRIZIRuf77PBJ/view?usp=sharing
• Cleaned up the routing within the website so it all connects
• Built the bookshelf that will be needed to hold the valves using wood scraps at the woodshop
• 3d printed the stir cup holder for the mixing of the drinks

 

Next Week’s Goals 

• Scale up the current one drink to have all 6 working by the demo
• Test the current implementation

Current Status: On Schedule

Personal Accomplishments

• Designed and built the drink holding structure using wood scraps in the wood shop. I intentionally used this as this is the cheapest option and allows us to have high structurally integrity. The structure is sturdy and doesn’t have any tilt or unbalance
• Integrated the moving wheel from last week with the solenoidal valves so they can happen at the same time
• Here is a video of it: https://drive.google.com/drive/u/0/home

Next Week Goals:

• Stress test the integrated valves and rotating wheel with various drinks
• Integrate Howard’s robotic arm with the code

Current Status: On Schedule

Personal Accomplishments

This week our team completed some major milestones in our project. We got the robot fully assembled and got some movement out of it established from the computer. We also got the garnish wheel mechanism connected to the motor for it and assembled the parts for the solenoid valve pipes. We also completed the 11 page Design Review Document. For the garnish wheel selection and the solenoid valves, we have the circuit drawn out for it so all we have to do is receive all of the parts to actually test and assemble it.

There were no changes to the initial plan from this past week as we are still working through the assembly phase. As we plan ahead of the actually integration of the parts, we believe that it is best to each take on our own part and then integrate it afterwards. Howard will be working on the robotic arm, Arda will be working on the garnish wheel and Rohan will be working on the solenoid valves.

Given our progress on this week, we have set ourselves up to make the individual circuits, established by the Arduino. This will lead us to integrate the individual parts soon.

Currently, we believe we are on schedule. We had to make a slight deviation in our plans considering some of our parts didn’t arrive in time and. That being said we accomplished some later goals (such as the Facial Recognition) to even us out.

 

Free Response

Part A (Written by Arda):

The Bar Belt system addresses a global need for automated beverage precision. With the increasing demand for automation in home entertainment and hospitality, our project provides a scalable solution that minimizes labor costs and optimizes each drink’s efficiency. Since it is a cloud based infrastructure, this will allow for easy updates and remote operation, which makes it adaptable to various markets beyond Pittsburgh, including urban centers, homes and any hospitality venue.

This project can be generalized to anyone who has access to internet. Since the project will be entirely automated, the user won’t have to do anything except input the drink that they would like to drink. The WebApp interface will be intuitive and clear so that anyone will be able to use it, engaging users with any level of computer expertise. By being scalable, sustainable, and accessible, Bar Belt is made to serve for a global market.

Part B (Written by Howard):

Our project is designed to be able to adapt to different cultural and social drinking traditions. For instance, while some cultures may emphasize the base ingredients used, some others may be more focused on getting a good ratio. To accommodate to these different approaches, our system allows for customized drink recipes, ingredient selections, and precise volume mixing, ensuring users can recreate drinks consistently that suit whichever cultural custom they have.

Additionally, our system can be expanded to non-alcoholic beverages and ingredients, expanding the scope for the amount of drinks that could be created and reaching a wider range of users who may not want alcoholic contents. By integrating a flexible recipe database and ingredient selection process, users can enjoy a variety of mixed drinks with ingredients that they are able to pick themselves. Our approach ensures that the system is not only functional but also culturally adaptable, making it suitable for a diverse range of social settings.

Part C: (Written by Rohan)

Our project is designed with environmental sustainability in mind by minimizing waste, conserving resources, and promoting eco-friendly practices. By precisely controlling liquid dispensing through solenoid valves, the system reduces spills and ingredient waste, ensuring efficient use of beverages and garnishes. The automated mixing mechanism eliminates the need for disposable stirrers, reducing plastic waste.

Personal Accomplishments

• Wrote the entire Design Document with the appropriate figures attached
• Built the parts for the garnish wheel selector with the motor attached
• Designed the circuit involving the motor

Next Week Deliverables

• Gather all of the parts needed for the circuit (including potentiometer, capacitors, motor driver)
• Have the circuit working with the motor moving

Status Report: On Schedule