3d Field evaluation warnings at high load

The resource consumption of the field evaluation directly depends on the number of measured points which lie in a field of an active evaluation. Under certain conditions, too many points in active fields can lead to overload of the sensor. This page describes conditions where overload warnings are issued and explains the limitations of these overload warnings. Furthermore, practical field configurations and common pitfalls with respect to the overload warnings are shown.

There are two different kinds of warnings which can displayed in the UI if there is the risk of overload due to the configuration of the field evaluation.

W1: "It is possible to activate evaluations with fields with too many intersecting beams"

W2: "Active evaluations have fields with too many intersecting beams"

To understand the difference between these two warnings, it is important to understand the following concepts

• Evaluations that can be activated are those evaluations which are currently not evaluated but which can be activated during runtime either via input or via telegram, depending on the configuration of the evaluation group.
• Active evaluations are those evaluations which are currently evaluated. Only those evaluations contribute to the resource consumption of the device.
• Evaluations which are configured but belong to a group which is deactivated do not require significant computation resources and therefore do not contribute to the computation of the warning criteria.

Knowing these concepts, one can conclude the following:

• W1 refers to a combination of evaluations which can potentially be activated. That means that the condition to trigger W1 is determined independently of the currently active evaluations. The implementation iterates over all (theoretically) possible activation combinations and checks if there is at least one which fulfils the warning conditions.
• W2 refers to currently active evaluations.

Both types of warnings can occur if either a certain number of evaluations is exceeded or a certain number of beams is exceeded which intersect the fields belonging to the evaluations. 

• For W1, all evaluation combinations which can be activated during runtime are taken into account.
• For W2, all currently active evaluations are taken into account.

The device offers a resource monitor which shows for each evaluation its share on the beam limit from the table above. It is important to note that this share does not reflect the share of computational resources which are actually used by this evaluation. There are two reasons for that:

1. Even if the resource monitor shows a resource requirement of 100% for all evaluations, that does not mean that these evaluations require 100% of the CPU. 100% means that no more evaluation should be activated in order to avoid data loss if the device is operated under full load (i.e. with streaming and all available filters switched on).
2. The criterion for the resource requirements is the number of beams that intersect a field. However, beams which intersect a field where the corresponding measurement points on these beams do not lie in the field  (green point below) require by far less resources compared to beams which intersect the volume where the corresponding measurement points are inside the volume (orange point below). This is because only measurement points inside the volume are taken into account for the object size computation, which is computationally expensive.

The example below shows two fields. Both fields contribute similarly to the resource requirements according to the resource monitor, but the consumption of computational resources of the left field is much higher compared to the right field since the right field contains much fewer points.