Abstract Clifford neural networks have gained significant attention from researchers in recent years. These networks are specifically designed to analyze data containing diverse geometric components. However, despite their growing adoption, leveraging these networks to handle large datasets with intricate geometric properties remains a considerable challenge. This proposal aims to harness the potential of Clifford neural networks for analyzing data with geometric attributes. Specifically, our focus is on exploring and evaluating their capability to analyze and predict point clouds, a critical aspect in the field of autonomous vehicles. The primary challenge lies in identifying the optimal geometric layers to extract the most effective representations from data. This is particularly complex for data with geometric properties that are not straightforward to capture or interpret. Motivation and Goal Clifford neural networks have emerged as powerful tools for analyzing data with geometric properties, offering unique advantages in domains where such structures are critical. These networks are increasingly utilized in applications requiring precise geometric analysis, such as point cloud processing and autonomous systems. However, effectively applying Clifford neural networks to large datasets with intricate geometric attributes presents significant challenges. Clifford neural networks hold great promise for advancing the capabilities of autonomous vehicles by enabling more accurate analysis and prediction of point clouds. Point clouds are essential in representing the 3D environment, and their accurate processing is critical for autonomous driving systems to perceive and navigate their surroundings safely. Given the importance of geometric data in autonomous vehicle systems, the development of robust and efficient geometric layers to extract meaningful representations is vital. This project aims to address this challenge by investigating and optimizing the architectural components of Clifford neural networks for processing complex geometric data. Our goal is to evaluate and enhance the capacity of these networks to analyze and predict point clouds, ultimately contributing to safer and more efficient autonomous vehicle systems. In addition to advancing the theoretical understanding of Clifford neural networks, we aim to provide practical insights into their application in safety-critical domains. This proposal sets the stage for groundbreaking work in geometric data analysis, focusing on improving the efficiency and reliability of autonomous vehicle systems.