🤖 Smart Vacuum Robot Simulation

Concurrent Perception & Mapping System – SPL225 @ BGU

A modular simulation of an autonomous vacuum robot’s perception and mapping system, developed as part of the Systems Programming Course at Ben-Gurion University.
The system models real-time sensor fusion using multithreaded Java Microservices and a custom MessageBus framework.

---

🔧 Technologies Used

• Java 8 – Core implementation using threads, lambdas, and generics
• Custom MessageBus – Publish/subscribe, event routing & futures
• Multithreading & Synchronization – Thread-per-service model
• GSON – For JSON parsing of input sensor data
• JUnit – Test-driven development for core components
• Maven – Build automation and dependency resolution
• Docker-compatible – Tested in isolated environments

---

💡 Project Structure

bgu.spl.mics.application

• messages/
  Contains all system messages (Events and Broadcasts) used in the pub/sub framework:

  • DetectObjectsEvent.java
  • TrackedObjectsEvent.java
  • PoseEvent.java
  • TickBroadcast.java
  • CrashedBroadcast.java, TerminatedBroadcast.java

• objects/
  Data representations of sensors, tracked objects, the robot pose, and statistics:

  • Camera.java, LiDarWorkerTracker.java, GPSIMU.java
  • DetectedObject.java, TrackedObject.java, LandMark.java
  • StampedCloudPoints.java, StampedDetectedObjects.java
  • FusionSlam.java, Pose.java, CloudPoint.java, StatisticalFolder.java
  • STATUS.java, LiDarDataBase.java

• services/
  Microservice implementations:

  • CameraService.java
  • LiDarService.java
  • FusionSlamService.java

---

✨ Features

• Fully event-driven concurrent design
• Global system timer using TickBroadcast
• Real-time pose-aware mapping with coordinate transformation
• Supports multiple sensors with different frequencies
• Fault detection & emergency shutdown using broadcast messages
• Generates a final JSON output containing runtime statistics, detected landmarks, and crash report (if applicable)

---

🚀 How to Run

1. Build

mvn clean compile

2. Run Tests

mvn test

3. Run Application

To run the simulation, provide the path to a valid configuration JSON file:

mvn exec:java -Dexec.mainClass=bgu.spl.mics.application.GurionRockRunner -Dexec.args="/path/to/configuration_file.json"

---

📄 Output Description

Generates an output_file.json in the same directory as the configuration file, containing:

• Runtime statistics
• Final map (landmarks)
• Crash report (if applicable)

---

🧪 Testing

mvn test

Includes unit tests for core components:

• MessageBusImpl – Publish/subscribe mechanisms and future resolution
• FusionSlam – Landmark transformation and pose estimation
• Sensor services (CameraService, LiDarService) – Message handling and event generation

---

📁 Directory Structure

src/
└── main/
    └── java/
        └── bgu/
            └── spl/
                └── mics/
                    └── application/
                        ├── messages/
                        ├── objects/
                        └── services/

---

🔗 Example Configuration

For testing the system, several input sets are provided under the following folders:

• example_input/
• example_input_2/
• example_input_with_error/

Each folder contains a configuration_file.json and supporting sensor data (camera_data.json, lidar_data.json, pose_data.json, etc.).

---

📚 Course Information

• Course: SPL225 – Systems Programming Lab
• Institution: Ben-Gurion University of the Negev
• Year: 2025
• Environment: Linux CS Lab, Docker-compatible

---

🧑‍💻 Authors

Ben Kapon
Student at BGU
LinkedIn

Itay Shaul
Student at BGU
LinkedIn

---

📝 Important Note

This project was designed and tested on a Docker-compatible environment.
Ensure file paths and JSON formats are valid when running in local or CI environments.

This project demonstrates real-time sensor integration using advanced concurrent programming techniques in Java.