Scribd
Upload
20 views4 pages

Assignment AV

The assignment for MT 310 Autonomous Vehicle focuses on applying data analytics and machine learning to solve real-world problems in autonomous vehicles, emphasizing hands-on experience with automotive datasets. Students will engage in a structured project that includes problem definition, literature review, dataset selection, exploratory data analysis, model development, evaluation, and reporting. The evaluation criteria will assess clarity, data handling, analytical approach, result interpretation, and documentation, with a submission deadline of April 24, 2025.

Uploaded by

kumaryajju29
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
20 views4 pages

Assignment AV

The assignment for MT 310 Autonomous Vehicle focuses on applying data analytics and machine learning to solve real-world problems in autonomous vehicles, emphasizing hands-on experience with automotive datasets. Students will engage in a structured project that includes problem definition, literature review, dataset selection, exploratory data analysis, model development, evaluation, and reporting. The evaluation criteria will assess clarity, data handling, analytical approach, result interpretation, and documentation, with a submission deadline of April 24, 2025.

Uploaded by

kumaryajju29
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

MT 310 Autonomous Vehicle II sem 2024 - 25

Assignment
Data Analytics in Autonomous Vehicles
Objective

To apply data analytics, machine learning, and statistical methods to extract insights, build
models, and solve real-world problems related to autonomous vehicles.

Learning Outcomes for Students

 Gain hands-on experience with real-world datasets from the automotive domain.
 Understand the pipeline: data collection → preprocessing → modeling → evaluation.
 Learn how to build ML/DL models tailored to high-dimensional, time-sensitive data.
 Develop critical thinking about ethics, safety, and interpretability in AV analytics.

Data analytics in autonomous vehicles involves using statistical analysis, machine learning,
and AI to process and extract insights from huge volumes of data collected by self-driving
vehicles.

Types of Data Involved

 Sensor Data: LiDAR, radar, camera, GPS, IMU


 Vehicle Telemetry: Speed, braking, steering angle, battery level
 Traffic and Infrastructure: Traffic lights, signals, GPS maps
 Environmental Data: Weather, road conditions
 User Behavior: In-cabin actions, preferences, usage patterns

Area Description
Perception Analytics Analyzing sensor data to detect lanes, vehicles, pedestrians
Predictive Modeling Forecasting actions of other vehicles/pedestrians (path prediction)
Driving Behavior Analysis Detecting anomalies or aggressive driving patterns
Fleet Management Optimizing routing, usage, and maintenance schedules
Safety & Risk Assessment Predicting collision probabilities or critical scenarios
Energy Optimization Analyzing battery usage or regenerative braking efficiency in EV-AVs
Simulation Analytics Testing AV scenarios using simulation data from CARLA, LGSVL, or BeamNG

Popular Tools & Technologies

Category Tools/Platforms
Data Analysis Python, R, Pandas, NumPy, Matplotlib, Seaborn
ML/DL Frameworks Scikit-learn, TensorFlow, PyTorch, XGBoost
Big Data Spark, Kafka, Hadoop (for real-time telemetry)
Visualization Plotly, Tableau, Power BI
AV Simulation CARLA, SUMO, LGSVL, AirSim

Faculty of Science & technology(IcfaiTech), IFHE Hyderabad


MT 310 Autonomous Vehicle II sem 2024 - 25

Sensor Processing OpenCV, ROS, PCL (Point Cloud Library)

Project Ideas

1. Object Detection in Urban Traffic Scenes


o Use YOLO or SSD models to detect vehicles, pedestrians, and traffic signs from
dashcam footage.
2. Predicting Vehicle Behavior Using Time-Series Data
o Apply LSTM/RNNs to predict the trajectory of nearby vehicles based on past
movement.
3. Data-Driven Lane Departure Warning System
o Analyze telemetry and camera data to issue warnings when the vehicle drifts from
its lane.
4. Real-time Collision Risk Estimation
o Build a model that predicts potential crash events using sensor fusion data.
5. Route Optimization for Autonomous Taxis
o Analyze historical traffic and customer data to optimize dispatch and routing.
6. Energy Efficiency Analysis of AVs
o Evaluate how autonomous systems impact energy consumption patterns in electric
vehicles.

Project Structure

1. Problem Definition

 Clearly define the objective.


 Identify key questions or hypotheses.
 Example: “Can we predict potential collisions based on real-time sensor data?”

2. Literature Review

 Study recent papers on AV data analytics.


 Benchmark existing methods or models.
 Include a minimum of 5 credible sources (IEEE, Elsevier, arXiv, etc.)

3. Dataset Selection

 Choose one:
o Public datasets (e.g., KITTI, Waymo Open Dataset, NuScenes, BDD100K)
o Simulated data (CARLA, BeamNG)
o Synthetic/generated datasets

Ensure you perform data wrangling, cleaning, labeling, and normalization.

4. Exploratory Data Analysis (EDA)

 Summary statistics
 Data visualizations (plots, heatmaps, distributions)
 Outlier detection and handling

Faculty of Science & technology(IcfaiTech), IFHE Hyderabad


MT 310 Autonomous Vehicle II sem 2024 - 25

5. Model Development

 Choose appropriate techniques:


o Classification (e.g., object detection)
o Regression (e.g., speed prediction)
o Clustering (e.g., event segmentation)
o Time-series forecasting (e.g., path prediction)
 Tools: Python, Pandas, Scikit-learn, TensorFlow/PyTorch, OpenCV, Seaborn/Plotly

6. Model Evaluation

 Split into training/test sets


 Use metrics like Accuracy, RMSE, Precision/Recall, F1-score, AUC
 Cross-validation or hyperparameter tuning encouraged

7. Results & Insights

 Interpret and visualize the results


 Compare with benchmarks
 Discuss limitations and future work

8. Presentation & Reporting

 Deliverables:
o Final project report (max 10 pages)
o Slide deck (10–12 slides)
o code, README, results
o Optional demo video (2–3 mins)

Timeline Suggestion (2-week course project)

Days Task
1–2 Topic selection, problem definition
Literature review, dataset
3–4
collection
5–6 Data preprocessing & EDA
7–8 Model development
9–
Model evaluation & tuning
10
11 Report writing & visualization
12 Final presentation/demo

Faculty of Science & technology(IcfaiTech), IFHE Hyderabad


MT 310 Autonomous Vehicle II sem 2024 - 25

Evaluation Criteria

Criteria Weight
Problem Clarity & Research Depth 15%
Data Handling & Preprocessing 15%
Analytical Approach & Modeling 25%
Result Interpretation & Insights 20%
Documentation & Presentation 25%

Submission date 24.04.2025

Note: The Assignment is submitted in a group of maximum 3 students. Report should consist of all
the three students and expected that all are involved in the process of development of the
assignment.

Faculty of Science & technology(IcfaiTech), IFHE Hyderabad

You might also like