Request handling over UDP

This project consists of developing a Room Booking and Meeting Scheduler in which a server handles client request over UDP. The goal of the project is to allow several clients to schedule meetings while inviting other peers, giving them the choice to accept, reject, withdraw, and join these meetings. The server acts as a central entity in order to take proper actions upon different types of message it receives from clients. In fact, the server writes its meetings in two different rooms depending on their availability

Smart Thermostat

This project consists of developing a smart thermostat that optimizes the heating and the cooling system of the house, the program consists of a broker that acts a server and handles all the input sent from the “Management APP”, the “Smart Door Locker” and the “Connected Thermostat”. The goal of the project is to allow several users to enter their names and their prefered temperature into the Management APP so that, upon their entering into the house, the thermostat would adapt to their prefered temperature following certain assumptions.

Air Traffic Controller

The goal for the project involves the creation of an air traffic monitoring and control system, our work includes all the part from design, implementation, testing and analysis of a simplified ATC. The scope of design will be created by us, whereas the input and output will be designed by our team.

Rescue Drone UAV

• •Full Electrical and Circuit design of the drone
1. PCB designed for the power distribution board.
2. Choosing the components, testing and validating on eCalc.
3. Integrated wireless communication for the remote and the laptop.
• •Software and Hardware Integration of the AI part in Python.
• •Developed python scripts to make an attached gopro detect a square shape, identify the color and the alphanumerical value it holds, as well as an approximate calculation of the geotag location it is on with the user of its current flight path.
• •Performed a financial analysis on the cost expectancies for the project.
• •Responsible for integration and testing of PixHawk flight controller:
1. o Created a fail-safe and a fail proof design feature.
2. o Performed multiple calibration testing.
3. o Mission planning and Ground Control Station Operation

Dot Puzzle solver

Dot Puzzle is played on a nxn board game. On each position of the board, a wooden token is plaaced. Each token has 2 sides, a white dot and a black dot. Initially the dotsare placed in a random position, the goal being to find the smallest number of moves to bring the board to its goal configuration where all the dots are white.

To solve this puzzle, DFS, BFS and Algorithm A* have been performed compared and then used to identify the fastest time.

Language Classifier

This project explores how language models can be built to facilitate the correct classification of the language of a tweet based on its content. In conjunction with the language models, a Naive Bayes Classifier is used to calculate the probability of each tweet’s correspondence to a particular language. To be more precise, the Pistachios' team worked on classifying tweets into different languages using three language models; unigrams, bigrams and trigrams. These models were built from a training set of 18318 labeled samples. To test the accuracy of the final product, a set of 7000 labeled tweets were used.

Intelligent Agent

Intelligent-Agent can answer university-related questions using a knowledge graph and natural language processing.

A webscraper was used to take all the courses in Concordia University including their name, code and description. After that the information is used to turn into triples using DBpedia topics. Finally Queuries are created so that the chatbot uses them to answer questions.