Fully automated driving

Bosch Automated Mobility Academy | Course 4

With fully automated driving, the car can manage all driving situations with no driver involvement whatsoever.

In this course, the car will learn how to fully utilize its technology and knowledge (software) to handle any and all driving situations.

Upon successful completion of this course, the vehicle will have achieved the ultimate goal: fully automated driving. The car's sensors, software, and artificial intelligence allow it to attain full certification, meeting stringent standards and regulations for performance, dependability, redundancy and safety. This is proof of performance, allowing the passengers to fully trust and depend on the vehicle to manage all driving responsibilities.

The car will become the third living space, allowing the passengers to freely engage in more enjoyable and productive pursuits while all of the driving duties are handled by the car.

As more and more vehicles achieve this level, the likelihood of accidents — and therefore road injury or death — will be reduced to almost nil. Productivity and efficiency on roadways will increase, mobility will be available to those who don’t have it, urban centers and traffic management will be transformed for the better, new economies will emerge, and the use of fossil fuels and greenhouse gas emissions from vehicles will be reduced.

Lesson 4.1

Artificial intelligence (AI)

In order to handle the unpredictable, the car must be able to think, learn and make intelligent decisions.

What is artificial intelligence?


Artificial intelligence (AI) describes a process by which machines learn to learn. Or to put it differently: AI is how computer systems imitate human intelligence.

In the case of automated driving, the Bosch AI onboard computer is expected to be able to guide self-driving cars through all traffic situations, including complex or new scenarios.

How does it work?


Many cars already use Bosch sensors to monitor their surroundings. Using artificial intelligence, the car will also be able to interpret those readings to make predictions about the behavior of other road users. For example, an AI system in a car can recognize pedestrians on the sidewalk, calculate the probability of someone crossing the road, and, if necessary, initiate evasive maneuvers.

What about the ethical and regulatory considerations of AI?


With cars taking on the driving and decision-making responsibilities, it's inevitable that difficult decisions will sometimes have to be made without utilizing human intuition. This raises many legal and ethical questions that industry partners are working together with regulators to address.


Lesson 4.2

Preparedness is everything

A fully automated vehicle must be prepared for all conditions, even technology failures and cyber security threats.

What happens if a critical component fails? What happens if hackers attempt to hijack the car?

These are very serious questions that must be answered in ordered to bring fully automated vehicles to the market — and bring peace-of-mind to users. Fortunately, Bosch is on it.

As vehicles gradually take over more and more driving tasks, safety-critical systems such as brakes and steering have to satisfy special requirements. For example, systems will be designed with redundancy, enabling the vehicle to bring itself either to a safe stop (i.e., pull over to the shoulder) or even continue to safely operate (i.e., by using back-up systems or "limp home modes") in the unlikely event of component failure.


A need for redundancy

Unlike today’s architecture where the driver is the backup in case of a failure (fail safe), future architectures with the driver no longer serving in that role still need to retain a certain level of backup performance (fail operational). Redundancy enhances the safety not only of the occupants, but also the vehicle and its surroundings.

To ensure this backup performance in case of a failure, the sensors, communication, power net, and actuators all need to be considered.

screen_shot_2017_05_11_at_3.24.42_pm_1 screen_shot_2017_05_11_at_3.24.42_pm_1_ background_trans


A fail-degraded brake system (and steering system and EE architecture) is required for highly automated driving to reach a safe state under all failure conditions (high deceleration and steerability required). One viable solution for a fail-degraded brake system is the combination of iBooster and ESP®. Both the iBooster and ESP® systems are independently capable of performing braking functions for the vehicle in the rare case of a single failure.


The vehicle must remain capable of steering in the highly unlikely event of component failure. The Bosch Electric Power Steeing system with fail-operational functionality enables either a driver or auto pilot system to make a safe stop in the rare case of a single failure.

Braking, steering and beyond

We tend to think first of braking and steering when it comes to redundancy, but there are several other important vehicle systems that also require special back-up systems.

This includes such systems as vehicle perception, localization and planning, information and display instruments, and – last but not least – an intelligent, well-planned E/E architecture to make it all possible.

More about redundancy for automated driving ›


Cyber security

A fully automated vehicle utilizes a variety of connected features, and in the course of doing its job, may communicate with other vehicles as well as infrastructure. This is called V2X communication.

Because of the connected nature of automated vehicles, as well as the need to protect data generated and used by automated vehicles, strong cyber security is critical.

This means that automated vehicles must not only contain on-board defense mechanisms to protect the vehicle and its systems from cyber attacks at the time the vehicle leaves the assembly line, but also be able to adapt to the changing threat environment years later. This can be achieved through any number of methods, including over-the-air updates, a technology Bosch is working hard to improve. Together with ETAS, its fully-owned subsidiary, Bosch is continuously developing cutting-edge security technology to shield vehicles from external threats while allowing necessary updates to pass through, ensuring the vehicle remains protected and ready to drive.

Lesson 4.3

Vehicle Localization: So the vehicle knows precisely where it is at all times. Bosch combines surround sensors, satellite navigation and inertial sensors to provide ultra-precise and reliable localization.

In the future, highly automated vehicles will be on the road in ever-increasing numbers. One important precondition for highly autonomous vehicles is precise localization. This means the vehicle must always know its current location with absolute reliability within a few centimeters.

GPS is a revolutionary tool that helps us get from A to B, but it is not sufficient for highly automated driving. The Bosch vehicle motion and position sensor (VMPS) enables vehicle localization through position signals from the Global Navigation Satellite System (GNSS), correction data and information from various sensors to determine the vehicles absolute position. This positioning ensures precise localization on a map and will enable highly automated driving.

Operation principle of localization

Bosch utilizes data collected by the vehicle’s surround sensors to create an independent map layer which depicts the unique features of a road – referred to as road signature. Advanced satellite-based localization, surround sensors and road signature are used to meet the high safety requirements of highly automated driving. The Vehicle motion and position sensor (VMPS) from Bosch plays a key role in this regard. It compiles information from high-performance satellite receivers, correction services for improving satellite positioning, inertial sensors, and intelligent software for reliable position calculation. The inertial sensors are used when environmental or surrounding information is insufficient or the satellite connection is malfunctioning or interrupted. In this case, the system calculates the vehicle’s relative position change by using sensor information and maintains the localization. VMPS is fully compliant to automotive standards and ISO26262.

Lesson 4.4

The physical transformation

No pedals, no steering wheel. This means BIG changes. The car itself must be able to handle all driving scenarios. With no active driver, this means that the vehicle itself transforms and truly becomes a third living space.


We believe there will be multiple, purpose-oriented interiors to maximize the mobility user’s newfound free-time. Imagine a seamless transition from home life and work life through the third living space, the convergence of home and work life on wheels. The autonomous car becomes a conference room, dining room, bedroom, hair salon, or movie theatre, the options are endless. New interior possibilities include the ability to rotate the front seats enabling all occupants to face one another, utilizing side windows as displays to project content, or adding a multi-touch table to enable collaborative work.

More about human machine interface

Application 4.1

The vision realized


At this stage, the car has now achieved the ability to operate in all situations completely without a human driver. As more and more fully autonomous vehicles hit the roads, the benefits will multiply. Accidents, traffic jams and road rage will be virtually unheard of. The way we as a society utilize mobility and vehicles will change. Infrastructures and whole cities will be reinvented to optimize the flow of goods and people, and we will usher in the era of the fully automated and connected car.

Congratulations, you're through Course 4!