oneM2M_Hackathon_Team_UmBokDong_2024

Topic

In many countries, reports of bicycle theft commonly occur. However, these problems are still not solved, because traditional bicycle locks used to slow down or merely prevent theft. Also, once a bicycle has been stolen, it is very difficult to get back from the one who stole it. This results in financial loss and inconvenience. We summarized the lack of functions and problems in traditional bicycle lockers.

• Absence of real-time monitoring
• Delayed theft detection
• Hard to retrieve stolen bicycles
• Inefficient manual locking system

To solve these kinds of problems, we represent our concept of a smart bicycle locker.

Our bicycle locker has two key functions.

First, when the device is turned on, it monitors the collision detection state and sends the collision state to the cloud server. If a user registers the device, the cloud service platform will send a collision detection message using MQTT Protocol.

Second, when the device is locked, Geo-fencing service will be activated and monitor its current location. If the device gets out of virtual boundaries, IN-CSE will judge that the device is being stolen and send a message to IN-AE. (This is one of the common service functions in oneM2M standards.)

Use Case of our project

1. Initial Setup

The user locks the bicycle remotely through the smartphone application. Upon locking, the following features are activated

Actuator Activation

• Bicycle Lock: The motor physically locks the bicycle.
• Geo-Fencing Setup: A geofence is created on the cloud based on the current bicycle location.

Sensor Activation

• GPS Module: Periodically updates the bicycle's location to the cloud.
• Collision Detection Sensor: Reports to the cloud if an impact above a certain threshold is detected.

2. Anomaly Detection

To seek anomaly, android applications will be connected with MQTT protocol in the background continuously.

Event 1: Exiting the Geofence Boundary

If the bicycle moves outside the set geofence

• The cloud triggers the bicycle's buzzer.
• Sends a notification to the user's smartphone.
• Bicycle location is constantly updated
• camera is activated to monitor the situation and surroundings.

Event 2: Collision Detection

If an abnormal impact is detected

• The buzzer is activated
• Sends a notification to the user's smartphone.
• Bicycle location is constantly updated
• camera is activated to monitor the situation and surroundings.

3. Response to Theft

The user tracks the real-time location of the bicycle through the application. If a user requests GPS data, the application will show the current and anomaly detected location on google map.

Remote Buzzer Activation

• Allows the user to identify the exact location of the bicycle through sound near its vicinity.

Remote LED Control

• Enables the user to turn the LED on or off for better visibility in low-light conditions.

4. Normal Usage

If the user controls the sensor, the application will send a HTTP message to the IN-CSE, and will execute appropriate commands.

When the user remotely unlocks the locker

• Unlocks: The motor disengages the lock.
• Deactivates Geofencing.
• The GPS module and collision detection sensor switch to sleep mode.
• To synchronize data between IN-CSE and IN-AE, IN-AE will request data from IN-CSE local database.

Architecture

1. Bicycle (Sensors and Actuators)

The bicycle is equipped with various IoT sensors and actuators, each performing specific roles while exchanging data with the cloud and the user application.

Sensors

• GPS Module(UART-GPS-NEO-6M): Activates when the bicycle is in a locked state. Sends the bicycle's current location to the cloud (Mobius).
• Collision Detection Sensor(SW-420): Activates when the bicycle is locked. Detects impacts above a certain threshold and notifies the cloud using the MQTT protocol.
• Camera Module(Raspberry Pi Camera 3): Activates upon user request via the application or when an anomaly is detected, capturing images and sending them to the cloud.

Actuators

• Lock(FIT0624 Inclined Electromagnetic Lock-12V): Engages or disengages the lock based on the user’s remote commands.
• LED Light: Can be turned on or off remotely by the user for safe nighttime riding or to locate the bicycle.
• Buzzer: Emits a warning sound when theft is detected or helps the user locate the bicycle.

2. Cloud Platform (Mobius)

Mobius serves as the core cloud platform for the project, managing data transfer and event handling between sensors and the user application.

Data Management

• Collects data from sensors via HTTP protocol.
• Stores the received data in Mobius' Server Database, making it available for retrieval upon user request.
• Updates real-time data such as Collision status using MQTT Protocol.

Subscription and Notification

• Sends messages to the users through subscriptions when sensor events (e.g., collision detection, exiting the geofence) occur.
• Ensures real-time response using MQTT-based notification messages.

Location Policy

• Activates geofencing based on user settings. Triggers warnings and activates the buzzer if the bicycle exits the designated boundary.
• Executes appropriate response processes based on location data and detected events.

Security and Access Control

• Protects access to sensor and application data through Access Control Policies (ACP).
• Ensures data synchronization and integrity while transmitting via secure communication channels (HTTP, MQTT).

3. User Application

The user application functions as an interface for interacting with the bicycle and the cloud, providing real-time information and control options.

Features

• Controls bicycle locking/unlocking, LED lights, and the buzzer.
• Displays real-time GPS location data integrated with Google Maps.
• Sends notifications for theft or impact detection to enable quick response.
• Allows users to view images captured by the camera in case of anomalies.

Operation

• Maintains a continuous connection with Mobius using the MQTT protocol to receive real-time alerts.
• Updates and controls sensor and actuator states through HTTP requests.
• Enables users to view data or send control commands via the application’s UI.

4. Resource Tree

Actual Implementation

ASN-AE & ASN-CSE

ASN-AE & ASN-CSE will be simulated by a device simulator program.This includes a GPS sensor, collision sensor, buzzer, led and motor (for lock mechanism).ASN-AE will create random data and send it to IN-CSE. Collision sensor data will be sent using MQTT Protocol. The rest of it will be sent using HTTP Protocol.

Sensors Activation

Actual sensors are activated by HIGH and LOW signals from the Raspberry Pi board. In our device simulator, we defined HIGH as 1 and LOW as 0.

1. GPS Sensor (UART-GPS-NEO-6M)

The GPS sensor transmits latitude and longitude in the DDDMM.MMMM format. However, we convert this to the DD.DDDD format for easier processing. This conversion is achieved by the formula: DD.DDDD = DDD + (MM.MMMM / 60).

2. Collision Sensor (SW-420)

The SW-420 collision sensor cannot directly measure collision intensity. Its sensitivity can be adjusted using a potentiometer. In our simulator, we modeled collision detection on a 0-10 scale, with the sensor activating when the value reaches or exceeds 8.

3. Buzzer (Passive Speaker Buzzer Module for Raspberry Pi)

The Passive Speaker Buzzer Module for Raspberry Pi actuator makes a loud and noise sound when the collision sensor is triggered. Also, it can be activated when IN-AE changed its state.

IN-CSE

We’ll use the Mobius, cloud service platform, for IN-CSE. IN-CSE will receive data from the device and store the data. IN-CSE will send subscribed data to IN-AEs using notification uris’ in the collision sensor data container.

IN-AE

We’ll use the android platform for IN-AE. ASN-AE registered in IN-CSE should record data in IN-AE using Room Database which is a local database of IN-AE. And, IN-AE will fetch the latest collision state using MQTT Protocol from IN-CSE and change attributes in the local database to use the fetched data for showing notification to user. Also, IN-AE can control led-light, lock-state, and buzzer by changing attributes in IN-CSE with HTTP Protocol. Finally, IN-AE will retrieve the ASN-AEs location data every 5 seconds.

Links

• Presentation Table
• Github Page