IBM Cognitive Coffee Machine

The design of CASE was achieved using electromechanical hardware controlled by a microprocessor, which needed to be connected to the Internet of Things (IoT) in order to link the hardware to IBM’s custom Application Programming Interfaces (APIs). These would allow the user to control the coffee machine by harnessing the power of artificial intelligence, using features such as voice commands and facial recognition. The project can be subdivided into three sections: electrical hardware, software and mechanical hardware.

The software was using IBM’s Bluemix and the Node-RED application as well as the Watson Conversation service in order to develop functionalities such as facial recognition, speech-to-text, interactive chat, full integration with our website/GUI and ability to control the general-purpose inputs and outputs (GPIOs) of the Raspberry Pi. Node-RED was used to link together different aspects of the project and bridge the gap between computer languages

The Mechanical hardware was designed using Computer-animated design (CAD). The case was a key task in the project as it holds the Raspberry Pi, Arduino, touchscreen and all the electro-mechanical hardware. The manufacturing of the machine was done through the combined use of 3-D printed parts and laser cut acrylic.

Finally the part I worked on was the electrical hardware. It allows the user to interact and control the coffee machine remotely without understanding how the machine itself works, acting as the bridge between software and mechanical hardware. It automate actions such as adjusting preferences or checking the water tank. We decided that CASE should be completely non-invasive, to make the product more reusable and versatile. The components are controlled by an arduino board who connects to the circuit and the raspberry pi directly as we can see below:

The firmware running on the microcontroller is very simple and uses a polling algorithm. It is self explained in the following diagram:

The project was supported by a website where our superviser could follow our progress and listen to our podcasts. The final product in action can be seen right here:

Project supervisor: Dr. Christos Papavassilou, Senior lecturer at IC.

IBM supervisor: Mr. John McNamara, Senior Inventor and Innovation Centre Technologist Lead at IBM.