Be sure to listen to this week's update on CASE Wednesday!

​

The third week of the project continued the progress that was made in the first 2 weeks. After getting most of the individual components of the project working, this week we worked on integrating different parts together to start building the final product.

 

The Hardware team and Software team started to collaborate and try getting some of the software components linked and working with the hardware. This was achieved by connecting the Raspberry Pi and Arduino via cables to the required output and input [ins, respectively. It was found that the Raspberry Pi can output PWM signals, which the Arduino could then use as an input signal. Thus after 3D printing improved parts for turning the dials and pressing buttons, these were connected to servo motors and then connected to the Arduino for testing. It quickly became apparent that there was not enough power to allow the use of all 5 servo motors that would be needed. To overcome this a power supply was used as a temporary solution until a better solution was found to be implemented into the final product.

The most important aspect, however, was using an output PWM signal from the Raspberry Pi to get the motors to move accordingly. This was tested by using an online form on the website to send an order that would produce an output from the Raspberry Pi that could then be used to provide the Arduino with the necessary input signal to move the motors accordingly. Using the external power supply, the required controls were successfully implemented by the Arduino to reflect the order made on the online form.

 

The Software team are still working on the functionality of the voice and facial recognition features. To improve response times, we are trying to continuously transcribe what the person is saying. Thus the machine would be taking in words as they are spoken rather than waiting for the end of a sentence and then using the speech to text API and determining what was said. For the facial recognition, we are working on correctly identifying the people that the camera sees, rather than just detecting that somebody is present. We are hoping that by next week we can start implementing these processes with the hardware as well to progress with the final product.

 

The Chatbot was also improved to incorporate facts about coffee and to provide more conversation rather than just taking commands for what kind of coffee a user would like. The structure of how coffee is order was also improved. Now the strength of the coffee is implied by the type of coffee, and the amount for an Americano has three distinct settings rather than the initially thought of 5 options. This was decided through testing the difference in taste when there were 5 possible strengths and discovering that adjacent levels had almost no difference in taste. With three possible settings in the amount of coffee, there are now distinct differences in the taste.

 

Progress has been made on the Graphical User Interface (GUI) that will be implemented in the final product. There is an options screen which shows the possible kinds of coffee that can be made, the time during standby, and the settings and progress screens are still being developed. The group also discussed possible case designs and decided for each member to come up with a couple sketches and then a final design would be chosen. The final design would allow for the Arduino, Raspberry Pi and necessary components to be securely held but also offer a sleek exoskeleton around the coffee machine. A draft of a case has been started on CAD and will be updated according to the final design chosen.

 

The group took photos for the About Us page, and are deciding on a suitable logo to use for the product. Currently, the name that has been decided for the product is CASE Coffee. This, however, may still change as the project develops.

 

After another productive week of work on the SMART Coffee machine project, the groups has made tremendous progress. We have managed to find a way of connecting the Raspberry Pi and Arduino to beginning integrating the software and hardware parts together. Further progress has been made on the facial and voice recognition services that will be used in the final product. A GUI is being developed to provide a nice interface on the touchscreen for the finals project. A case has started to be designed and will be one of the main focuses for the next week. Overall the group is enjoying the project, and are working hard to make a quality final product.