How to transmit Android real-time sensor data to computer?

Question

I wish to transmit real-time sensor data collected by Android smartphone to my computer and do the signal process on my computer. How may I achieve that? Any helpful links to tutorials are very well welcomed.

Either by wireless means or USB cables is acceptable.

When the data are transmitted, how may the computer process them?

I am familiar with Python, and so perferrably use Python to deal with the coming data.

Is it possible for Python to continuously accept newly come data and process them?

Answer 1

You could use python sockets to recieve the data and process it. Look at: https://docs.python.org/2/howto/sockets.html for some idea of how to setup the server machine.

Android has a compatible socket that you could use to send your data: Look at: http://examples.javacodegeeks.com/android/core/socket-core/android-socket-example/ for some ideas on how to use it.

Edit: This will work for wifi and 3g/4g.

Answer 2

There exist multiple Android apps to transmit sensor data wirelessly. Checkout e.g. "Smartphone IMU GPS" [1], which is open-source [2]. It transmits the data via UDP. A Python code snippet to receive the data stream is given in the app description [1]. Make sure the smartphone as well as the receiving device are connected to the same WiFi network and no firewall blocks the traffic.

[1] https://play.google.com/store/apps/details?id=de.lorenz_fenster.sensorstreamgps
[2] https://sourceforge.net/projects/smartphone-imu/

Answer 3

Some Android apps allow you to share the sensors via the network:

• PhonePi and SensorServer (using websockets)
• Sensor Node (via MQTT)
• IP Webcam (can be used as a webapi by accessing http://SMARTPHONE_IP:8080/sensors.json)
• Tasker publisher

You can also read the sensors via ADB!

Answer 4

You can use Sensor Server app from Github, which streams live sensor data to WebSocket clients.

To receive live sensor data from Android , you simply need to connect to the app using following URL

ws://ip:port/sensor/connect?type=<sensor-type>

where <sensor-type> is the type of sensor you are connecting to. For example

For Accelerometer : /sensor/connect?type=android.sensor.accelerometer

For orientation : /sensor/connect?type=android.sensor.orientation

For gyroscope : /sensor/connect?type=android.sensor.gyroscope

and so on...

Almost every language provides implementation of Websocket protocol. To receive live data in Python script you can use WebSocket client for Python

import websocket



def on_message(ws, message):
    print(message) # sensor data here in JSON format

def on_error(ws, error):
    print("### error ###")
    print(error)

def on_close(ws, close_code, reason):
    print("### closed ###")
    print("close code : ", close_code)
    print("reason : ", reason  )

def on_open(ws):
    print("connection opened")
    

if __name__ == "__main__":
    ws = websocket.WebSocketApp("ws://192.168.0.102:8082/sensor/connect?type=android.sensor.accelerometer",
                              on_open=on_open,
                              on_message=on_message,
                              on_error=on_error,
                              on_close=on_close)

    ws.run_forever()

Sensor Server app and Websocket client must be connected to the same network

Answer 5

I was trying to get real time audio data from my phone to python. The apps mentioned in the other answers can't read the microphone, but phyphox can.

First you choose what you want to measure in phypox, I'll take raw audio as an example, but it probably works for other sensors too and audio types too.

First go to audio autocorrelation and go to the tab raw data. When you press play you can see a graph of what will be received on the pc. Then select "allow remote access" from the menu in the top right. A url appears in at the bottom of the screen.

The following code can be run on your pc and will display a graph of an instantaneous measurement, but it can of course be adjusted to continually receive data:

import requests
import json
import matplotlib.pyplot as plt


def send_request():

    try:
        response = requests.get(
            url=f"{url}/get?recording=full&time=full"       # this can be changed for other audio types
        )
        print(f'Response HTTP Status Code: {response.status_code}')
        # print(f'Response HTTP Response Body: {response.content}')
        return response.content
    except requests.exceptions.RequestException:
        print('HTTP Request failed')


def graph(raw):
    data = json.loads(raw.decode('utf-8'))              # Turns data into a large dictionary
    amplitude = data["buffer"]["recording"]["buffer"]   # This location depends on the sensor type
    t = data["buffer"]["time"]["buffer"]
    print(amplitude)
    plt.plot(t, amplitude)
    plt.show()


url = "url"         # insert url from phyphox here

raw_data = send_request()
if raw_data:
    graph(raw_data)

When looking at other sensors, the full url for requests.get() can be found by opening the url that phyphox gives in your browser, then use inspect element, and under the tab "network" a lot of data is coming in. Click on one of these and the link will show up at the top of the headers.