All Updates

  • Friday, Feb 16, 2024
blog-image

3rd Quarter Updates , 15/02 - 08/04

  • In the first week of the semester, we estabilished all of our periodic meetings with our coordinators throughout the entire period;

  • The framework hugo was chosen as the tool with which our website was going to be created. The theme for the website was picked so that it could work as a landing page for the project and as a blog;

  • The work was split between all team members;

  • Our team developed the first version of our prototype’s block diagram, which can be found here :Block Diagram.

  • A first version of the electronic connections of our system has been developed , which can be found here : Eletronic layout.

  • We discussed the use of RFID tags in our project, mainly its advantages and disavantages in relation to the project’s goals.

  • The first Electrocap seminar took place on 21/02, where the topic of “Design thinking” was presented. In this presentation we had the opportunity to listen to the advices given by the speaker, a former IST student , about risk minimization in products and companies, what to look for when inovating a product or a service and how to structure a team capable of completing the tasks related to project developing.

  • The team was able to implement a Python program that can detect whether a person enters or leaves a specified room. The program was tested only using one person.

  • The second Electrocap seminar took place on 06/03. The theme of this seminar was “The Power of Teamwork”, where the speaker presented his experience in team based projects, as well as some advices on how to organize tasks between team members, set up meetings and how to communicate effectively.

  • The use of RFID technology was disencouraged given the possible problems in terms of reliability (detection of tags, distance from the RFID reader);

  • We’ve tried to contact the company Prosegur in order to get to know their opinion on modern smart alarm systems and our approach to the problem of Smart Home Security (sitll no answer as 29/03/2024);

  • In preparation for the ElectroCap mid-program pitch deck, our team has designed the block diagram corresponding to our system’s connections and operability.

  • After abandoning the RFID technology, our team developed a new schematic for our prototype, which can be found here: Access Controller Schematic. This new version uses the same equipment, but with the addition of a magnetic contact.

  • The third Electrocap seminar took place on 18/03. The theme of this seminar was “The art of prototyping”, where the speakers talked about their experiences regarding prototype design, implementation an testing.

4th Quarter Updates : 14/04 - 31/05

  • The fouth Electrocap seminar took place on 17/04. The theme of this seminar was Funding, where the speaker spoke about his experince on the field of startups,as well as investment and funding opportunities.

  • The team planned and divided the work for the remaining quarter of classes. The work was split into three categories: Application development, where the main goal is to build an app to serve as a user interface for our automated alarm system, allowing the system to effectively communicate with the user; MQTT protocol implementation (Message Queuing Telemetry Transport),where the objective is to program the broker and subscriber sides of the communication between the app (the user) and the system. This specific protocol was chosen since it is widely used in IoT communication and it is easily implemented (it is TCP based). The third task is Raspberry pi tests, where the facial recognition algorithm will be tested for further optimization and changes, as well as the program that coupled with an overhead camera can detect people entering a specific room.

  • When testing our facial recognition algorithm, we had to reframe the images captured by the camera , as well as the capturing rate, since the limited computational power and memory (2 GB RAM in this Raspberry Pi model) was not enough for the peak of memory reached during the facial recognition. The solution our team decided was to resize the captured images and to downsample the capturing rate to 10 fps.

  • Our team developed an initial version of the Android App with the framework Flutter, using the Dart programming language. The layout of the application was tested, as well as the the communication between the App (running in a computer) and another device, in which an image was sent to the device running the app, which was successfull.

  • We began testing the components of our prototype individually, in order to prevent malfunctions when building the final version. THE PIR sensor and the magnetic contact were tested successfully. We performed various tests in both cameras (in case of the USB camera, we tried different resolutions and frame acquisition rate(fps) in order to prevent the Raspberry Pi from crashing). We started developing the main code, where all the different parts of our project will be integrated.

  • In the weekly meeting with our coordinators, at 13/05, we performed our first demonstration of the facial recognition algorithm with the PIR sensor.

  • Our team configurated the Firebase server that will host different accounts for our mobile application. Our system sends a message to a Flask server, which routs the message to the Firebase Cloud Messaging , which sends the notification to the user.

  • An example of our mobile application receiving notifications from our system after the detection of an authorized and/or unauthorized entrance can be found here: Notifications through the AI-LARM APP.