📱 Mobile APP Mapping and Navigation

1. Course Content

• Master using the [ROSRobot] mobile APP on mobile terminals (Android / iOS) to wirelessly remote control the A1 Pro robot for real-time environment mapping, map saving and loading.

• Master receiving real-time video stream, observing occupancy raster map and laser point cloud, calibrating initial pose via touch screen, and sending navigation goal points for autonomous cruising on the mobile APP.

2. Preparation

1. Install APP: Search and install ROSRobot in the mobile app store (or scan the QR code in the manual to download and install).

2. LAN Connection: Ensure the phone and robot are connected to the same WiFi LAN (or connect to the robot's hotspot WiFi), and record the robot's current IP address (can be viewed by executing ifconfig in terminal).

3. Mobile APP Mapping Operation

3.1 Start Camera Video Streaming Node

Based on the camera type used, start the corresponding camera node in Terminal 1:

xxxxxxxxxx
# If using depth camera:
ros2 launch yahboomcar_depth camera_app.launch.py

# If using ordinary USB monocular / robotic arm camera:
ros2 launch usb_cam camera.launch.py

3.2 Start APP Mapping Service (Choose One)

Run in Terminal 2 the APP mapping bridge node (this Launch automatically pulls up chassis driver, lidar, SLAM mapping node, Rosbridge WebSocket bridge, and APP map save service):

xxxxxxxxxx
# Method A: Gmapping algorithm mapping (recommended)
ros2 launch yahboomcar_nav map_gmapping_app_launch.xml

# Method B: Cartographer algorithm mapping
ros2 launch yahboomcar_nav map_cartographer_app_launch.xml

3.3 APP Connection and Mapping Operation

1. Connect to Robot: Open the ROSRobot APP on the phone, select mapping mode. Enter the robot's LAN IP address, keep the port as default 9090, VideoTopic select /usb_cam/image_raw/compressed (if using Nuwa camera, select the corresponding compressed topic), and click 【Connect】.

   

2. Remote Control Mapping: After successful connection, the APP interface will display the real-time camera video feed and the raster map generated by the lidar. Use the virtual joystick in the lower-left corner to remote control the robot to move slowly and smoothly, gradually scanning and exploring the entire indoor environment.

   

   

3. Save Map: When the environment outline is completely closed, click the 【Save Map】 button in the upper-right corner of the APP interface, enter the map name (for example test) in the popup, and click confirm.

   

4. View Saved Maps: The APP map save service will automatically call map_saver_cli in the background to save the map file to the robot system path:

   xxxxxxxxxx
   ls ~/yahboomcar_ros2_ws/yahboomcar_ws/src/yahboomcar_nav/maps/
   

   

4. Mobile APP Autonomous Navigation Operation

4.1 Start Chassis and Lidar

Run in Terminal 1 (start A1 Pro chassis driver and lidar):

xxxxxxxxxx
ros2 launch yahboomcar_nav laser_bringup_launch.py

4.2 Start Camera Video Streaming Node

Run in Terminal 2 the camera node:

xxxxxxxxxx
# Nuwa camera:
ros2 launch yahboomcar_depth camera_app.launch.py

# USB camera:
ros2 launch usb_cam camera.launch.py 

4.3 Start APP Navigation Service

Run in Terminal 3 the navigation node:

xxxxxxxxxx
#map:=The path needs to be modified according to the actual mainboard
ros2 launch yahboomcar_nav navigation_teb_app_launch.xml map:=/home/jetson/yahboomcar_ros2_ws/yahboomcar_ws/src/yahboomcar_nav/maps/test.yaml

4.4 APP Navigation Operation

1. Connect to Robot: Open ROSRobot APP, select navigation mode, enter robot IP address and port 9090, VideoTopic select /usb_cam/image_raw/compressed, and click 【Connect】.

   

2. Enter Navigation Main Interface: After successful connection, the APP automatically loads the map and robot pose:

   

3. Set Initial Point (2D Pose Estimate): Click 【Set Initial Point】, click at the actual physical position of the robot on the phone's map screen and pull out the chassis heading arrow. Observe that the red/green scanning point cloud on the map coincides with the black obstacle contours, representing successful initial pose alignment.

   

4. Set Navigation Goal Point (2D Goal Pose): Click 【Set Navigation Point】, click in the traversable free area on the map and set the goal heading. The robot immediately plans an optimal global path, autonomously controls steering and forward movement, smoothly avoids obstacles, and arrives at the destination.

   

5. Core Source Code and Communication Configuration Analysis

5.1 APP Communication Bridge Architecture Analysis

The mobile APP achieves full-duplex communication with the ROS 2 system through standard WebSocket protocol:

1. rosbridge_websocket:

   • Listens on port 9090, responsible for transparently forwarding ROS 2 topics, services, and mobile APP JSON messages in both directions.

   • APP-published control commands are transformed into /cmd_vel to drive the chassis; APP-set pose points are transformed into /initialpose and /goal_pose and passed to Nav2.

2. robot_pose_publisher:

   • Real-time listens to the map -> base_footprint in the TF transformation tree, calculates the robot's absolute coordinates and heading in the global map, and high-frequency sends to the APP interface via /robot_pose topic to draw the robot icon.

3. laserscan_to_point_publisher:

   • Converts lidar scan data into a compact point cloud format, providing real-time obstacle point cloud outline for APP rendering.

APP Navigation Launch Architecture Configuration (navigation_teb_app_launch.xml):

xxxxxxxxxx
<launch>
    <!-- Start rosbridge websocket bridge service -->
    <include file="$(find-pkg-share rosbridge_server)/launch/rosbridge_websocket_launch.xml"/>
    
    <!-- Start lidar point cloud publishing node -->
    <node name="laserscan_to_point_publisher" pkg="laserscan_to_point_publisher" exec="laserscan_to_point_publish"/>
    
    <!-- Start TEB navigation core -->
    <include file="$(find-pkg-share yahboomcar_nav)/launch/nav2/navigation_teb_launch.py"/>
    
    <!-- Start robot map pose publishing node -->
    <include file="$(find-pkg-share robot_pose_publisher_ros2)/launch/robot_pose_publisher_launch.py"/>
</launch>

5.2 APP Map Save Service Source Analysis

Source Code Path:

xxxxxxxxxx
$HOME/yahboomcar_ros2_ws/yahboomcar_ws/src/yahboomcar_app_save_map/yahboomcar_app_save_map/yahboom_app_save_map.py

xxxxxxxxxx
class MinimalService(Node):

    def __init__(self):
        super().__init__('minimal_service')
        # Create ROS 2 service to receive APP map save requests
        self.srv = self.create_service(WebSaveMap, 'yahboomAppSaveMap', self.add_three_ints_callback)
        
    def run_shellcommand(self, *args):
        '''Execute shell command and run map_saver_cli in background'''
        args = sum([(arg if type(arg) == list else [arg]) for arg in args], [])
        possess = subprocess.Popen(args, stdout=subprocess.PIPE)
        return possess

    def add_three_ints_callback(self, request, response):
        map_name = request.mapname
        workspace_src = Path.home() / 'yahboomcar_ros2_ws' / 'yahboomcar_ws' / 'src'
        map_path = str(workspace_src / 'yahboomcar_nav' / 'maps' / map_name)
        database_path = workspace_src / 'yahboomcar_app_save_map' / 'yahboomcar_app_save_map' / 'xgo.db'
        
        # Record map name and timestamp information into local SQLite database
        now = datetime.datetime.now()
        str_time = now.strftime("%Y-%m-%d %H:%M:%S.%f")
        map_id = hashlib.md5((str_time + map_name).encode()).hexdigest()
        response.response = request.mapname
        
        try:
            conn = sqlite3.connect(str(database_path))
            c = conn.cursor()
            c.execute("INSERT INTO xgo_map (map_name, map_id, map_path) VALUES (?, ?, ?)", (map_name, map_id, map_path))
            
            # Call Nav2 official map_saver_cli node to save .yaml and .pgm raster map
            self.run_shellcommand(
                'ros2', 'run', 'nav2_map_server', 'map_saver_cli',
                '-f', map_path,
                '--ros-args',
                '-p', 'save_map_timeout:=10.0',
                '-p', 'map_subscribe_transient_local:=false',
            )
        except sqlite3.IntegrityError as e:
            response.response = {"msg": str(e)}
            return response     

        self.get_logger().info('Map saved successfully: %s -> %s' % (map_name, map_path))
        return response

• This node creates the yahboomAppSaveMap service (type is WebSaveMap). When the APP clicks Save Map, it sends a request with the map name; while writing map metadata to SQLite database, the node asynchronously calls nav2_map_server's map_saver_cli command-line tool to persistently save the current global occupancy raster map as maps/<map_name>.yaml and maps/<map_name>.pgm.

5.3 Core Communication Parameters and Port Description