Introductory description

The aim is to help students build a solid knowledge of key AI techniques that are widely used in development of autonomous vehicles. To understand the application context, an overview of autonomous vehicles technology will be provided, including: Localization, Sensing & Perception and Motion Planning. The module will then focus on practical aspects of AI while the students will gain a strong high level understanding of the underlying theory. The emphasis will be on Deep Learning techniques that are heavily used in driverless vehicles, including: Convolutional Neural Networks (CNN), Supervised and Unsupervised Learning, Recurrent Neural Networks.

Module aims

Outline syllabus

This is an indicative module outline only to give an indication of the sort of topics that may be covered. Actual sessions held may differ.

• An overview of autonomous vehicles technology
• System architecture
• Localisation
• Sensing and perception
• Motion planning in complex environments
• A general overview of AI systems
• Data science basis for machine intelligence:
• Understanding experimental data and fitting
• Clustering and classification
• Deep learning systems
• Introduction to neural networks
• Deep learning neural networks
• Reinforced learning
• Supervised and unsupervised learning
• Convolutional neural networks
• Recurrent neural networks
• Industry expert seminars
• Tutorials on tools and examples

Learning outcomes

By the end of the module, students should be able to:

• 1. Demonstrate a complete understanding of the high level architectures and principles of autonomous driving systems
• 2. Demonstrate systematic high-level knowledge of the AI systems
• 3. Demonstrate a critical understanding of neural networks architecture
• 4. Demonstrate a comprehensive understanding of Deep Learning
• 5. Demonstrate the mastery of relevant tools used to achieve machine intelligence
• 6. Critically evaluate the socio economic implications of the use of AI technologies in automated transport
• 7. Practically analyse and optimise different types of neural networks used for different automotive tasks
• 8. Critically analyse data sets and techniques for clustering and classification

Indicative reading list

• GOODFELLOW, Ian; BENGIO, Yoshua; COURVILLE, Aaron. Deep learning (adaptive
  computation and machine learning series). Adaptive Computation and Machine Learning
  series, 2016, 800.
• RUSSELL, Stuart Jonathan, et al. Artificial intelligence: a modern approach. Upper Saddle
  River: Prentice hall, 2003.
• URMSON, Chris, et al. Tartan racing: A multi-modal approach to the darpa urban
  challenge. 2007.
• GUTTAG, John V. Introduction to computation and programming using Python. Mit Press,

• SAMARASINGHE, Sandhya. Neural networks for applied sciences and engineering: from
  fundamentals to complex pattern recognition. CRC Press, 2016.
• ASIMOV, Isaac. I, Robot, Robot series. 1950.

A variety of up-to-date sources including:

• Latest government / UK Automotive Council roadmaps for autonomous vehicles
• Latest automotive legislation and standards
• Current academic research in the field of smart connected autonomous vehicles

Subject specific skills

Basic knowledge of AI techniques that are widely used in development of autonomous vehicles, Overview of autonomous vehicles technology: Localization, Sensing & Perception and Motion Planning, Deep Learning techniques that are heavily used in driverless vehicles, including: Convolutional Neural Networks (CNN), Supervised and Unsupervised Learning, Recurrent Neural Networks.

Transferable skills

Critical Thinking, Problem solving, Communication, Information literacy (research skills), Digital literacy, Professionalism