Hierarchy Of The Driving Task To Design ADAS

The three-level hierarchy of the driving task according to Donges (1982) is a model that describes the different cognitive demands of driving. The three levels are:

1. Navigation level: This level is concerned with the planning of the route and the selection of the appropriate speed. It is the highest level of the hierarchy and requires the most cognitive effort.
2. Command level: This level is concerned with the definition of a trajectory for the vehicle. It takes into account the current situation, such as the traffic conditions and the road layout, and generates a plan for how to reach the destination.
3. Stabilization level: This level is concerned with the control of the vehicle to follow the desired trajectory. It is the lowest level of the hierarchy and requires the least cognitive effort.

The three levels of the hierarchy are not always clearly separated. For example, the driver may need to make decisions about the route at the command level, or they may need to make adjustments to the trajectory at the stabilization level. However, the hierarchy provides a useful way to think about the different cognitive demands of driving.

The three levels of the driving task are hierarchical, meaning that the higher levels depend on the lower levels. For example, the command level cannot function without the navigation level providing it with a route to follow. Similarly, the stabilization level cannot function without the command level providing it with a trajectory to follow.

The three-level hierarchy of the driving task can be used to understand the different types of cognitive demands that are placed on drivers. For example, navigating a new city is a knowledge-based task that requires the driver to use their knowledge of the road network to plan a route. On the other hand, driving on a familiar route is a skill-based task that requires the driver to rely on their experience to make smooth and safe lane changes.

The three-level hierarchy of the driving task can also be used to design automated driving systems. For example, an automated driving system that is only capable of performing the stabilization level of the task would be able to control the vehicle's lateral and longitudinal motion, but it would not be able to plan a route or make decisions about the best way to get to the destination. An automated driving system that is capable of performing all three levels of the task would be able to drive the vehicle from start to finish without any input from the driver.

Here is a table that summarizes the three levels of the hierarchy:

Level	Description	Cognitive Demand
Navigation	Planning the route and selecting the appropriate speed	High
Command Guidance	Defining a trajectory for the vehicle	Medium
Stabilization	Controlling the vehicle to follow the desired trajectory Controlling the vehicle's lateral and longitudinal motion	Low

The time horizons typical for the task levels of navigation, guidance, and stabilization are as follows:

• Navigation: The navigation task level is responsible for planning and executing a path from the vehicle's current location to its destination. The time horizon for navigation is typically on the order of minutes to hours.
• Guidance: The guidance task level is responsible for keeping the vehicle on its planned path. The time horizon for guidance is typically on the order of seconds to minutes.
• Stabilization: The stabilization task level is responsible for keeping the vehicle's attitude and heading constant. The time horizon for stabilization is typically on the order of milliseconds to seconds.

The following table provides a more detailed breakdown of the time horizons for each task level:

Task Level	Time Horizon
Navigation	Minutes to hours
Guidance	Seconds to minutes
Stabilization	Milliseconds to seconds

It is important to note that these are just general guidelines. The actual time horizons for each task level will depend on a number of factors, such as the vehicle's speed, the complexity of the environment, and the level of automation desired.

For example, a vehicle traveling at high speed will need to have a shorter navigation time horizon in order to avoid collisions. A vehicle operating in a complex environment, such as a city, will also need to have shorter time horizons for all three task levels in order to react to changing conditions.

The level of automation desired will also affect the time horizons. A vehicle with full automation will need to have shorter time horizons in order to ensure safety. A vehicle with partial automation, on the other hand, may be able to have longer time horizons, as the driver can take over in case of an emergency.

The time horizons for each task level are also affected by the accuracy requirements of the application. For example, an application that requires high accuracy will need to have longer time horizons than an application that does not require as much accuracy.