Live Telemetry Charts From BeamNG.drive using Python
Overview
BeamNG.drive is described as a “dynamic soft-body physics vehicle simulator capable of doing just about anything,” and prides itself on its accurate physics engine and extremely detailed damage model. This allows users to explore and drive whatever they wish in BeamNG.drive with its various sandbox levels and scenarios. Usually, BeamNG.drive is recognised for its damage model, and indeed, there are many videos on YouTube showcasing cars crashing and being damaged in various ways. However, if we set aside the damage and crashes, we see a simulator that offers a very realistic driving experience. Every car component is simulated and can be adjusted/tuned to match personal driving styles. This, combined with the damage model, provides a real sense of risk and immersion when driving in the game.
I bought BeamNG.drive sometime in 2020 and didn’t play it extensively, occasionally loading it up and driving around, crashing various cars. Very recently, I bought a Thrustmaster racing wheel to upgrade from my Logitech wheel, and since then, I have been playing it a great deal. I downloaded a map for the Nürburgring-Nordschleife and have been lapping it in a GT3-esque car, constantly tuning and adjusting to get a car that fits my driving style. It got me thinking if there was any way that I could pull telemetry data from BeamNG.drive and plot it, to see if it would reveal any areas of the lap I could improve on.
BeamNG.drive and the OutGauge Protocol:
After conducting some research, it was discovered that BeamNG.drive utilises the OutGauge protocol to send out UDP packets of data that can be received and unpacked. I decided to write a program in Python that would receive the data, unpack it, and then plot live graphs. The code for the program is provided below, accompanied by explanations of each section.
Code Breakdown
Plotting imports
This section simply imports the needed libraries to be able to run the program properly
from itertools import count
import pandas as pd
import matplotlib.pyplot as plt
from matplotlib.animation import FuncAnimation
Create UDP socket
To be able to receive the UDP packets, a socket had to be created and then bound to the IP address and port as specified in BeamNG.drive
import socket
import struct
sock = socket.socket(socket.AF_INET, socket.SOCK_DGRAM)
# Bind to BeamNG OutGauge.
sock.bind(('127.0.0.1', 4444))
sock.setblocking(False)
Receive UDP data and unpack
I defined a function called receiveData() that would, initially, have no UDP data and therefore assign a variable data to None, to indicate this lack of data. While data == None, the receiveData() would try to pull data from the UDP packets BeamNG.drive was sending. If successful, receiveData() would then unpack the incoming data and assign more useful variables such as gear, speed, RPM, etc. Finally, receiveData() would exit and return the required values for the graphs.
def receiveData():
data = None
while True:
try:
data, fromAddr = sock.recvfrom(4096)
except socket.error:
break
if data is None:
return None, None, None, None, None, None
outsim_pack = struct.unpack("I4sH2c7f2I3f16s16si", data)
gear = outsim_pack[3]
speed = 2.23694 * outsim_pack[5]
rpm = outsim_pack[6]
engTemp = outsim_pack[8]
fuel = 100 * outsim_pack[9]
throttle = 100 * outsim_pack[14]
brake = 100 * outsim_pack[15]
clutch = 100 * outsim_pack[16]
# print(f"RPM: {rpm:4.0f} Speed: {speed:3.0f}")
return throttle, brake, clutch, rpm, speed, fuel
Real-time graph
Since the data points on the axes of the graph would constantly be changing, [] had to be left empty to allow the data would be assigned to specific points along the graphs.
plt.style.use('fivethirtyeight')
x_vals = []
y_throttle = []
y_brake = []
y_clutch = []
y_rpm = []
y_speed = []
y_fuel = []
This short section simply positions 4 sub-plots within 1 figure, to allow 4 real-time graphs to be displayed
# Settings for subplots (Total number of plots, column position, row position)
fig = plt.figure(figsize=(12, 12))
ax1 = fig.add_subplot(4, 1, 1)
ax2 = fig.add_subplot(4, 1, 2)
ax3 = fig.add_subplot(4, 1, 3)
ax4 = fig.add_subplot(4, 1, 4)
A function animate(i) had to be made to allow the graphs to display real-time data. Assigning throttle, brake, etc… To receiveData() allows animate(i) to pull the returned data from earlier and make use of it within the graphs. Next, the returned data is appended to various axes each update, such that the graphs constantly update with new incoming data.
def animate(i):
throttle, brake, clutch, rpm, speed, fuel = receiveData()
x_vals.append(next(index))
y_throttle.append(throttle)
y_brake.append(brake)
y_clutch.append(clutch)
y_rpm.append(rpm)
y_speed.append(speed)
y_fuel.append(fuel)
The next portion of code sets up the graph axes, limits, line weight and colour, then updates the graph every 10ms.
# Needed to stop graphs changing colours
ax1.clear()
ax2.clear()
ax3.clear()
ax4.clear()
# Moving x axis
if len(x_vals) > 100:
ax1.set_xlim(x_vals[-100], x_vals[-1])
ax2.set_xlim(x_vals[-100], x_vals[-1])
ax3.set_xlim(x_vals[-100], x_vals[-1])
ax4.set_xlim(x_vals[-100], x_vals[-1])
# Setting vertical axis limits
ax1.set_ylim(-5, 105)
ax2.set_ylim(-150, 10000)
ax3.set_ylim(-5, 180)
ax4.set_ylim(-5, 105)
# Plotting graphs, with labels and colours
ax1.plot(x_vals, y_throttle, lw=1, color='blue')
ax1.plot(x_vals, y_brake, lw=1, color='red')
ax1.plot(x_vals, y_clutch, lw=1, color='green')
ax2.plot(x_vals, y_rpm, lw=1, color='black')
ax3.plot(x_vals, y_speed, lw=1, color='black')
ax4.plot(x_vals, y_fuel, lw=1, color='black')
ax1.set_ylabel('Throttle, brake, clutch (%)')
ax2.set_ylabel('RPM')
ax3.set_ylabel('Speed (MPH)')
ax4.set_ylabel('Fuel (%)')
plt.tight_layout()
index = count()
ani = FuncAnimation(fig, animate, interval=10)
plt.show()
Program Demonstration
The video below shows a demonstration of the program plotting my inputs for the throttle, brake and clutch (blue, red and green lines respectively), along with the RPM, speed and fuel level of the current vehicle.
Final Words
While the program runs as intended, there are limitations that can be worked on in the future. The main limitation currently is the update rate of the graphs. The program can plot just a single graph very smoothly; however, when more than one graph is required, the program has to plot these simultaneously, leading to slower (but still tolerable) graph plotting.
Full Code
# Plotting imports
from itertools import count
import pandas as pd
import matplotlib.pyplot as plt
from matplotlib.animation import FuncAnimation
# UDP set up
import socket
import struct
# Create UDP socket.
sock = socket.socket(socket.AF_INET, socket.SOCK_DGRAM)
# Bind to BeamNG OutGauge.
sock.bind(("127.0.0.1", 4444))
sock.setblocking(False)
# Receive UDP data and unpack
def receiveData():
data = None
while True:
try:
data, fromAddr = sock.recvfrom(4096)
except socket.error:
break
if data is None:
return None, None, None, None, None, None
outsim_pack = struct.unpack("I4sH2c7f2I3f16s16si", data)
gear = outsim_pack[3]
speed = 2.23694 * outsim_pack[5]
rpm = outsim_pack[6]
engTemp = outsim_pack[8]
fuel = 100 * outsim_pack[9]
throttle = 100 * outsim_pack[14]
brake = 100 * outsim_pack[15]
clutch = 100 * outsim_pack[16]
# print(f"RPM: {rpm:4.0f} Speed: {speed:3.0f}")
return throttle, brake, clutch, rpm, speed, fuel
# Real-time graph
plt.style.use("fivethirtyeight")
x_vals = []
y_throttle = []
y_brake = []
y_clutch = []
y_rpm = []
y_speed = []
y_fuel = []
# Settings for subplots (Total number of plots, column position, row position)
fig = plt.figure(figsize=(12, 12))
ax1 = fig.add_subplot(4, 1, 1)
ax2 = fig.add_subplot(4, 1, 2)
ax3 = fig.add_subplot(4, 1, 3)
ax4 = fig.add_subplot(4, 1, 4)
def animate(i):
throttle, brake, clutch, rpm, speed, fuel = receiveData()
x_vals.append(next(index))
y_throttle.append(throttle)
y_brake.append(brake)
y_clutch.append(clutch)
y_rpm.append(rpm)
y_speed.append(speed)
y_fuel.append(fuel)
# Needed to stop graphs changing colours
ax1.clear()
ax2.clear()
ax3.clear()
ax4.clear()
# Moving x axis
if len(x_vals) > 100:
ax1.set_xlim(x_vals[-100], x_vals[-1])
ax2.set_xlim(x_vals[-100], x_vals[-1])
ax3.set_xlim(x_vals[-100], x_vals[-1])
ax4.set_xlim(x_vals[-100], x_vals[-1])
# Setting vertical axis limits
ax1.set_ylim(-5, 105)
ax2.set_ylim(-150, 10000)
ax3.set_ylim(-5, 180)
ax4.set_ylim(-5, 105)
# Plotting graphs, with labels and colours
ax1.plot(x_vals, y_throttle, lw=1, color="blue")
ax1.plot(x_vals, y_brake, lw=1, color="red")
ax1.plot(x_vals, y_clutch, lw=1, color="green")
ax2.plot(x_vals, y_rpm, lw=1, color="black")
ax3.plot(x_vals, y_speed, lw=1, color="black")
ax4.plot(x_vals, y_fuel, lw=1, color="black")
ax1.set_ylabel("Throttle, brake, clutch (%)")
ax2.set_ylabel("RPM")
ax3.set_ylabel("Speed (MPH)")
ax4.set_ylabel("Fuel (%)")
plt.tight_layout()
index = count()
ani = FuncAnimation(fig, animate, interval=10)
plt.show()
# Release the socket.
sock.close()