How to Check Vacuum Cliff Sensor FOV in CAD

Robotic vacuum cleaners rely on optical cliff sensors to detect drops and prevent falls down stairs. In Computer-Aided Design (CAD) environments, engineers validate the field of view (FOV) of these sensors to ensure zero blind spots and eliminate physical obstructions caused by the chassis, wheels, or bumper. This article details the CAD modeling techniques, geometric simulations, and kinematic analyses used to verify cliff sensor placement and optical coverage prior to physical prototyping.

1. Optical Cone Generation

Engineers begin by modeling the optical characteristics of the infrared (IR) emitter and phototransistor receiver as 3D solid or surface geometry:

  • Emitter and Receiver Cones: Standard CAD packages (such as SolidWorks, CATIA, or Siemens NX) allow users to extrude conical volumes representing the sensor's angular emission pattern and receiver acceptance angle.
  • Angular Limits: These cones are defined using manufacturer datasheets, typically ranging between 15° and 60°, depending on the chosen lens and sensor package.
  • Triangulation Modeling: Cliff sensors function via optical triangulation or time-of-flight (ToF). Engineers model both the projection cone and the reflected detection cone to identify the precise overlap zone (the active sensing volume) at standard operating floor clearances (typically 10 mm to 20 mm).

2. Interference and Clearance Checking

Once the optical cones are modeled, interference detection tools confirm the line of sight:

  • Chassis Shadowing: CAD software runs boolean clearance checks to confirm that the sensor cones do not intersect with neighboring hardware, such as side brushes, caster wheels, or the edge of the bumper.
  • Recessed Window Protection: Cliff sensors are commonly recessed within protective optical windows to avoid scratching. The CAD model verifies that the cone clears the bezel edges of the recess without internal reflection or attenuation.

3. Edge and Drop-Off Geometry Simulation

To evaluate actual fall detection, engineers place the CAD assembly into virtual test environments that replicate floors, stair ledges, and dark surfaces:

  • Threshold Detection Planes: A virtual ground plane is placed beneath the robot. By progressively offsetting or terminating this plane to simulate an edge, engineers evaluate at what distance the active sensing volume loses contact with the floor.
  • Approach Angles: The robot assembly is rotated and translated toward the edge at varying angles (perpendicular, 45-degree angle, reverse) to calculate the exact trigger point relative to the wheel contact patch.

4. Dynamic Kinematic Analysis

Static CAD checks are insufficient because vacuum cleaners tilt, pitch, and bounce during travel:

  • Suspension Travel Verification: Engineers articulate the CAD model's drive wheel suspension through its full stroke (compression and rebound) to ensure the FOV continues to hit the floor at an acceptable angle of incidence.
  • Bumper Compression: Kinematic simulations verify that when the front bumper is depressed against an obstacle, it does not swing downward into the sensor's FOV.
  • Transition Thresholds: Moving the model across transitions, such as rug edges or door thresholds, verifies that harmless height changes do not prematurely enter the sensor's "cliff" trigger zone.

5. Automated Ray Tracing and Optical Add-ins

For advanced validation, mechanical models are paired with optical simulation plug-ins (such as SPEOS or FRED):

  • Ray Tracing: Thousands of virtual optical rays are fired from the emitter to map exact ground reflection intensities back to the receiver.
  • Material Mapping: Engineers assign optical surface properties (such as black absorption, matte finish, or high-gloss tile) to the floor geometry to verify signal return strength across varying floor types.