What's the Difference between Autonomous, Automated, Connected, and Cooperative Driving?

Even nearly everything has already been said by anybody in the topic of automated driving, the basic terms are still mixed and misused by imprecise differentiation. Therefore, we try another attempt in the basic definition here and add some arguments why one needs to be more precise in the terminology:

Autonomous Driving: Autonomy means, that one is able and allowed to make decisions indepentendly and on one's own mind. In the case of autonomous driving a single vehicle can make its own driving decision independently.

Automated Driving: Automation deals about the execution of processes and procedures without human intervention. Hence automated driving implies driving without the intervention of human drivers. Sometimes a further differentiation between fully automated driving and the different automation levels according SAE is done.

Connected Driving: In the case of connected driving information is exchanged between automated as well as non-automated vehicles and other traffic participants and/or infrastructure in an automated way.

Cooperative Driving: Cooperative driving means, that single vehicles and drivers act cooperatively within traffic. This implies, that single traffic participants are coordinating their microscopic aims and actions in the light of improved overall macroscopic effects.

Automated Traffic: Automated traffic not only deals with the automation of single traffic participants but with the automation of traffic as a holistic system-of-systems. Hence the automation of traffic control and infrastructure is included in addition to the vehicle automation.

As a consequence of the above definitions some conclusions can be drawn:

  • Connected driving improves the abilities for cooperative driving, because traffic participants can express and share their intention more easily and precisely. This helps the other traffic participants for an improved consideration in their cooperative actions. Due to the fact, that the coordination with the help of connected driving is not only possible at microscopic level but also on macroscopic level, further potential for improvements in traffic efficiency and effectiveness are released.
  • Connected driving itself does not necessarily imply cooperative driving. Single traffic participants can theoretically use the additional information for their own individual advantage at the cost of others.
  • Autonomous driving does not intrinsically cause improved traffic. If everybody decides on his own without a cooperative coordination with other traffic participants, then chaos and traffic collapses may be a consequence. Normally autonomy is only appropriate in the case of low densities.
  • Automated driving can lead to significant improvements in traffic, because cooperative behavior can be enforced for robots much easier than for human beings. Robots follow their instructions much more precisely than humans, unless these are autonomous robots which decide to do differently.
  • Connected driving can accelerate the introduction of automated driving significantly, because it enables approaches for collective learning to identify and resolve inappropriate behavior and driving strategies quickly.
  • Automated driving often depends on valid decisions, which rely on information that can not be aquired with physical sensors. Such information often could be provided by connectivity.
  • If the term autonomy is taken seriously, then autonomous vehicles must also perform well without connectivity.
  • Cooperative and connected driving can lead to improved traffic even without automation.
  • Cooperation is also possible without connectivity in the case the (automated) traffic participants strictly follow given rules and conventions. But these rules then must be consistent, complete, and acceptable in context to the individual targets (which is far from being trivial!).
  • Autonomy normally leads to better acceptance, because of self-reliant decisions. But in case of saturated and crowded traffic situations in fact the improvements come from reduction of autonomy and increase in coordinated cooperation. Once a crowded traffic situation is cleared again, one can switch back to autonomy if the situation permits (according to the VERONET concept).

This means, that autonomy is not equivalent to automation and connectivity is not equivalent to cooperativeness.

Though autonomy may improve acceptance, but may lead to chaos and collapse in the case of high densities and saturation. Automation may improve cooperativeness, because robots normally do precisely what they are told to do. Connectivity enables an improved exchange of information for coordination of interests by active cooperation.

Isn´t this all here small-minded hairsplitting? Not at all! The implementation of vehicle and traffic automation is primarily based on performant software and "intelligent" control algorithms. For the specification and development of these a clear and well sorted mind is mandatory, when bad surprises are to be avoided. The expected advantages by vehicle and traffic automation have to be sorted out and developed very well, if positive effects in all target dimensions should be achieved at all (compare Balancing the benefits of automated driving). One of the mandatory requirements for automation is the clear formulation and specification of what has to be done in which way. A clear and crisp terminology is a must for this.


Last update on 2018-06-23 by Andreas Kuhn.

Go back