Use Case Examples for Rule Engine
In this page you can find examples of rules for real world use cases.
Last updated
In this page you can find examples of rules for real world use cases.
Last updated
Detect vehicles (motorized traffic) passing a street in the wrong direction (e.g.: one-way-streets of highway entrances).
As a first step, the scenario needs to be configured on camera level. Follow the setup guideline for a standard traffic counting use case. Create a new rule, name it and choose the configured counting line (CL). For wrong-way drivers, a predefined template can be used. You still have the opportunity to adapted it according to your needs. For the wrong-way driver you can create a rule that the direction needs to equal "out" which in your configured scene needs to be the wrong direction.
At an intersection, only detect objects which are performing a U-turn. As a first step, the scenario needs to be configured on camera level. Follow the setup guideline for a standard intersection monitoring use case.
Create a new rule, name it and choose Origin destination as triggers for the rule. For U-turns, a predefined template can be used. You still have the opportunity to adapted it according to your needs. Therefore, you can connect the existing origin and destination zones in your scenario and in case anyone is going from one zone back again to the same, one can assume, that this was a U-turn.
In traffic situations, there are several situations where a given class of street users should not use dedicated areas, e.g.:
people in the center of an intersection
Vehicles in fire-service zones
In order to check when and how often it happens, you can create a rule based on a predefined RoI in these dedicated areas. Create a new rule, name it and choose the RoI as triggers for the rule. You can find a template as an example for "person on street".
In the subcondition you can choose "Object" as a parameter and choose min nb of objects which need to apply to the conditions. You can define which classes are expected or not. On top, a dwell time condition can be added in order to only take objects into account which are in the area longer than a given time. (e.g. jaywalking, wrong parking in fire-service zones).
How often has it happened to any of you have been cut off at a pedestrian crossing while crossing or waiting to cross. This happens on a daily basis, and quite often it is very close to severe incidents. In order to know if and how often this happens, we provide you a solution with our rule engine. This is the basis for you to know where to set dedicated actions. The solution is a combined rule with a CL that is detecting the vehicles and a RoI which is focusing on pedestrians and bicycles. Configure a CL or speed line in front of the pedestrian crossing. On top, an RoI can be configured at the pedestrian crossing and/or the waiting area next to.
With that configuration, one or several rules can be created. In this example, one rule for this high risk situation is defined. You can detect when at least one person is on the Pedestrian crossing and a vehicle is crossing the speed line at more than 10 km/h.
Here is a short video where it is shown how such a rule will be applied.
Define at least one ROI and create an associated rule. As long as the rule is valid, the associated Quido relay output is enabled (contact closed). One or more rules can be created for the same ROI.
Please contact Support if you would like to try out this feature or if you have any further questions.
