CICLE-IA

Smart Module for sustainability and health promotion.

This work is now at the stage of embedded Machine Learning usage on open hardware platforms.
The research consists of managing power in a bicycle's electric drive-train system and biological data from the rider. We are prototyping different hardware model designs so they will be able to receive the Tensor Flow Lite platform, and that will allow us to run the machine learning models inside a microcontroller.
Such development will allow the improvement of the "Moove Verde" project, started in 2017 at "Labtec@" (our Free Hybrid Learning Environment) by researcher Deyvid Abreu, a Civil Engineer undergraduate student from IFAL - Campus Maceió. He is developing the topic "Sustainable Urban Mobility - Models and their feasibility for Maceió-AL" as his Capstone Project and under the guidance of Professor Marcelo de Assis Corrêa.

Participation on VII Ibero-American Congress on Entrepreneurship, Energy, Environment and Technology (CIEEMAT 2022)

The so-called "CICLE-IA" will integrated into the 2Moove platform and the Moobi Verde prototype, both developed between 2019 and 2020. These two together are a developed business model consisting of a virtual bike racks platform for sharing electric bicycles inside a determined location or campus through charging pile stations. The users will then be able to "rent" the bicycles and give them back at one of these stations.

The model involves a charging point station for charging the batteries through Electrovoltaic panels. The charging piles are connected to Base Radio-Stations allowing the interoperability of the model, or the IoT (Internet Of Things). Through a smartphone app, the users will dispose of different features regarding using electric bicycles as an urban transport method.

Prototype No. 1

The first step in this project was the prototype developed between 2018 and 2019. Back then, we implemented adaptations in a regular bicycle using low-cost components for the electric drive-train implementation. Using a DC (Direct Current) motor running at 48V powered by a set of 4 lead-acid batteries of 12V and conventional ratchets and chains, easily found in our national market, we docked a traction model at the bike pillion. To control the Voltage and Current, we implemented a microcontroller and a DC coupling relay.
Making this model showed us the feasibility of building a low-cost electric bicycle. Also, the tests and essays allowed us to expand our means to enhance power control involving the battery charging system.

This picture shows a prototype from mid-2019, and to validate the built electric drive-train model we did a bunch of tests on it. Carrying out the tests and making the prototype gave visibility to the project, which brought other people who may be interested in contributing and have a better understanding of our work proposal. With the new members, new solutions were prospected to solve our issues with the drive-train and battery charge/discharge. That was when the Moobi/2Moove proposal emerged.

Electric Charging Piles - The prototype

Between 2019-2020, involving technologies like IoT and geolocation the concept of a charging station using photovoltaic panels and an automatic charging controller for bicycles began to take shape. More students and researchers started to engage in the project. The subject "Sustainable Urban Mobility" expanded to improve the app, businesses' models, and technologies such as IoT, LoRa, and embedded intelligent systems.
The prototypes in the pictures on the side are part of the results obtained by 2020. Unfortunately, due to the pandemic situation, our work was focused only on developing the firmware and software, like the App MoobiVerde, and implementing the GPS/GSM for system management.

App MoobiVerde - Version 1.5

During the proposal phase of this business model emerged the idea of using geo-locators with a smartphone. For this, we developed an app using React, receiving collaboration from undergraduate students of Information Technology from IFAL - Campus Maceió.
The app MoobiVerde enables the following system functionalities:

Electric bicycle sharing inside a specific location or campus, where the user can rent one and give it back at one of the solar-powered stations.
Bicycle renting through the app and a QR code on it, unlocking its padlock.
Geolocation of the bicycle identifying its operational status.
Data storage of traveled distance log.
Performance and autonomy indicators.

This app is being updated to version 2.0, receiving new features for the forthcoming biological-data sensors that will integrate the MoobiVerde/2Moove system.