Continental Automotive | AI-Empowered Mobility

Facilitating the mobility of the future – with AI-enabled products and services

Passenger cars that independently search for and find a parking space a few blocks away. Commercial trucks that haul freight from Dallas to Los Angeles without a driver interfering. School buses that detect if a child has been left behind in the back row. This will be the new normal in the mobility of the future. And this future is already knocking on our door – with the help of innovative and reliable AI applications providing support, comfort and safety to drivers and passengers, in both private cars as well as commercial vehicles.

As an AI-empowered company, our aim is to use new AI technologies to develop the most cutting-edge portfolio of AI-enabled products, services, and methods that fulfill the needs of future mobility – and that will help unleash the vast potential this future has to offer. Our AI experts are currently working on exciting solutions for the AI-empowered mobility both in software as well as hardware.

New AI technologies will help us build software solutions on the basis of suitable hardware platforms, enhancing human capabilities"

Dr. Pia Dreiseitel

Head of AI Technologies at Continental

No Artificial Intelligence Technology Without a Powerful Hardware Platform

What are the crucial ingredients for AI technologies to make the mobility of the future possible? Data, software, and computing power. That is why, nowadays, software is the starting point for any vehicle development. You can learn more about the Software-Defined Vehicle here and scroll down to some examples of how we use AI-empowered software to improve some of our own solutions. But without a powerful hardware platform and without a vehicle architecture that can handle the different functions, autonomous driving will never be possible. The key lies in forming strong partnerships where we can work on both sides of the AI coin together.

Bringing Generative AI Into The Vehicle

In 2023, we have formed a partnership with Google Cloud to integrate its data and Artificial Intelligence technologies directly into vehicle computers. By being able to use generative AI, drivers will have a unique and intuitive experience interacting directly with their car.

Smart Cockpit HPC Google Cloud CES2024.jpg

The base for this collaboration is Continental’s Smart Cockpit High-Performance Computer (HPC) that combines both user experience and system performance. By offering manufacturers a pre-integrated set of functions and extensive options for customization, it is the perfect hardware platform enabling autonomous driving functions at the interface between car and user.

Learn more about our partnership with Google Cloud

Making Autonomous Trucking Systems Possible

Autonomous driving is also the hope for commercial transport. It promises increased safety, reduced fuel consumption, and quicker delivery times. That is why Continental has partnered up with Aurora to offer the industry's first hardware-as-a-service business model, based on mileage driven, through the Aurora Horizon platform.

Continental PP Aurora Quokka US-Truck Lowres

In this unique partnership, Continental provides the autonomous driving system kits including sensors, automated driving control units (ADCU), high performance computers (HPC) and telematics units as well as the complete fallback system which covers the full chain of effects. The first units will be available from 2027 onwards.

Learn more about autonomous trucking and our partnership with Aurora

Using AI to Improve and Develop Autonomous Driving Functions

In addition to providing sophisticated hardware and HPC solutions, we also use AI technology to improve existing functions and develop software for the software-defined vehicle. The following examples showcase some of the areas we are currently working on.

AI-Empowered Cameras and Surround Views

Perception and vision play an important role in Advanced Driver Assistance Systems (ADAS) already today.

However, for the autonomous car of the future to react appropriately in any situation and plan its further path, it needs to reliably detect and identify 3D objects, including their position, orientation and motion.

In our next generation cameras, AI technologies will calculate this optical flow rate in the camera itself which relieves the load on downstream processing units and reduces latencies by eliminating the need to transmit data.

Learn more about our camera technologies

Furthermore, current virtual camera applications used in low-speed maneuvering situations rely on highly distorted surround-view.

By adding AI-based approaches and apply AI technology to the surround-view camera images, we aim to render a virtual third person viewpoint that will give the driver a live view of the desired viewpoint and therefore better orientation.

Learn more about our human vision technologies

AI-Empowered Safety and Comfort

AI-empowered technology can also enhance the vehicle experience even if there is no driving involved. Continental’s Smart Access Solution – CoSmA – allows the driver to access and start the vehicle by using a digital key on a smart device.

Its ultra-wideband technology enables real-time precise positioning of the smart device and an advanced algorithm reliably detects the smart phone’s location inside or outside of the vehicle. Furthermore, CoSmA has sensing features that can detect a child left behind in the car and alerts the driver within seconds.

Learn more about our digital access solution CoSmA

AI Methods & Technologies

As an AI-empowered company, we apply various methods, tools, and technologies from the vast universe of Artificial Intelligence when developing hard- and software for future mobility – from Natural Language Processing to Time Series Prediction, from Enterprise Semantic Knowledge to Cognitive API Generation. Here are some examples how we work with and use them.

Uncertainty-based Detection

Blind spot monitoring system visually displayed, alerting the driver to unseen obstacles in urban traffic

Safety is paramount when it comes to autonomous driving functions, and this is closely related to how reliable the AI models are.

By using a method called Uncertainty-based Detection, we track the statistical uncertainties in Artificial Intelligence models which allows us to identify unusual patterns of behavior and gain more knowledge on how the model is working. Through this transparent approach, we help make the AI more reliable, robust, and accurate.

Data + Knowledge = Hybrid AI

Fully-autonomous driving requires not just the application of a single AI technology, but the combination of several different methods with data, rules, and human expert knowledge to form a holistic Hybrid AI system.

Concept Of Human Intelligence And Technology With Focus On Deepmachine Learning. Сoncept Human Intelligence, Technology, Deep Machine Learning

We apply the state-of-the-art research in Hybrid AI in our own use cases, e.g.

long-term traffic scenario prediction,
trajectory prediction,
representation learning from heterogenous sources or
explaining machine learning models and their plausibility.

Learn more about Hybrid AI in our PPP project “AI Knowledge”

Revolutionizing Perception using AI

Effective AI perception requires more than a single technology; it necessitates mastering various advanced methods, data sources, and expert knowledge into a cohesive system. Continental Engineering Services applies state-of-the-art research in AI to shape our Multipurpose Learning Network, tailored for diverse applications, such as:

Object detection and classification
Semantic segmentation
Skeleton pose estimation

The Multipurpose Learning Network is designed to handle input from multiple camera types, versatile for various operational environments, including on-road, off-road, railways, and agricultural settings on various platforms. Key features include:

Ready for Use: Optimized and pretrained neural network for immediate application
Pretrained: minimizing data acquisition and labeling efforts for specific applications.
Multi-Task Learning: Capable of managing multiple tasks with different objectives simultaneously reducing the process time and hardware resources.
Seamless Integration: Fast and Easy integration across various platforms,
No Copyright Concerns: All data used is available for commercial use.

Back to Artificial intelligence

AI-empowered Mobility