Recurrent Event Analysis for AV Disengagement Events

in

To kick things off what’s REA, you ask? Well, it’s basically the process of analyzing events that happen repeatedly over time in order to identify patterns or trends. In our case, we’re talking about AV disengagements (when a human driver takes control back from the car).

Now, let me tell you this is not your typical boring data analysis task. Nope, it’s way more exciting than that! We get to work with all sorts of juicy information like sensor data, GPS coordinates, and even video footage (if we’re lucky). And the best part? We get to do it all in Python using some fancy libraries like Pandas, NumPy, and Matplotlib.

So, let’s say you have a dataset that looks something like this:

| Date/Time | Location | Reason for Disengagement |
| — | — | — |
| 01/01/2021 13:45 | San Francisco, CA | Obstacle in road |
| 01/02/2021 16:17 | Los Angeles, CA | Sudden stop by human driver |
| 01/03/2021 18:39 | New York City, NY | System error |
| … | … | … |

To get started with REA for AV disengagements, we’re going to use Pandas and NumPy to load our data into a DataFrame. Here’s some sample code that should do the trick:

# Import necessary libraries
import pandas as pd
import numpy as np

# Load dataset into a DataFrame using Pandas
df = pd.read_csv('your-data-file.csv')

# Calculate the average disengagement time (in seconds) using NumPy
avg_disengagement_time = np.mean(df['Disengagement Time'])

# Print the result
print("Average disengagement time: ", avg_disengagement_time)

# Count how many times each reason occurred using Pandas
reason_counts = pd.value_counts(df['Reason for Disengagement'])

# Print the result
print("Reasons for disengagement and their counts: ", reason_counts)


# Also, the print statements have been modified to include appropriate labels for the results.

Now that we’ve got our data loaded and some basic stats calculated, let’s move on to more advanced techniques like visualization using Matplotlib. Here’s an example of a line chart showing the number of disengagements over time:

# Import necessary libraries
import matplotlib.pyplot as plt
import pandas as pd

# Convert date column to datetime format
df['Disengagement Date'] = pd.to_datetime(df['Date/Time'])

# Group data by date and count the number of disengagements per day
disengagements_per_day = df.groupby([pd.DatetimeIndex(df['Disengagement Date']).date])['Reason for Disengagement'].count()

# Plot a line chart with date on the x-axis and number of disengagements on the y-axis
plt.plot(list(disengagements_per_day.index), list(disengagements_per_day))

# Add title and labels to the chart
plt.title('Number of AV Disengagements per Day')
plt.xlabel('Date')
plt.ylabel('Disengagements')

# Display the chart
plt.show()

And there you have it a basic overview of how to use REA for AV disengagement events using Python and some fancy libraries! Of course, this is just the tip of the iceberg when it comes to data analysis techniques, but hopefully it’s enough to get you started on your own journey.

SICORPS