the camera on the front for obstacle avoidance seems useless, it just runs over my dog's toys and socks

Diagnosis

The Roborock Q Revo MaxV utilizes a sophisticated ReactiveAI 2.0 system that combines a structured-light 3D sensor with an RGB camera to identify objects. When the vacuum runs over socks or dog toys, it is typically due to a failure in "object classification." The system must not only detect a physical mass but correctly categorize it as an "obstacle" rather than a "surface variation." If the object is too small (like a thin sock) or lacks a distinct color contrast against your flooring, the AI may dismiss it as a rug edge or a floor shadow. Furthermore, the MaxV's vision is heavily dependent on ambient light. While it has a built-in LED light for dark areas, the light's throw is limited. If the robot is operating in a dim hallway or under a low-hanging table, the RGB camera cannot resolve the edges of a toy, causing the robot to rely solely on the bumper sensors. Once the bumper makes physical contact, the robot has already "run over" the item. You can confirm this by checking the "Obstacle Avoidance" log in the app to see if the robot tagged the item as a "small object" or ignored it entirely.

How to Fix It

To optimize the ReactiveAI 2.0 system for better detection of small household items, follow these precise steps:

1. Open the Roborock App and select your Q Revo MaxV. Navigate to Settings > Robot Settings > Obstacle Avoidance.
2. Ensure the ReactiveAI toggle is set to "On." If it is already on, toggle it off and back on to refresh the firmware handshake.
3. Locate the Collision Avoidance Mode setting. Switch this to "Standard" or "Avoid." If it is set to "Less Collision," the robot is programmed to prioritize cleaning coverage over avoidance, which leads it to "nudge" objects like socks.
4. Navigate to Settings > Robot Settings > Pet. Enable the "Pet" mode if available for your firmware version; this adjusts the sensitivity of the AI to recognize organic shapes and common pet accessories.
5. Physically inspect the front camera lens—the small circular glass aperture located just above the front bumper. Use a clean, dry microfiber cloth to wipe away any film, dust, or fingerprints. Even a light smudge can blur the RGB image, making a sock look like a flat part of the floor.
6. Perform a "Test Run" in a room with high lighting (open curtains) to determine if the issue is light-dependent.

If That Didn't Work

• Update Firmware Manually: Go to Settings > Firmware Update. If an update is pending, install it immediately. Roborock frequently pushes "Object Library" updates that improve the recognition of specific items like power cables and clothing.
• Establish "No-Go Zones" for High-Clutter Areas: If your dog's toys are concentrated in one area (e.g., a toy bin corner), use the Map editor to draw a "No-Go Zone." This prevents the robot from entering the area entirely, bypassing the need for AI detection in high-risk zones.
• Check for "Invisible" Obstacles: If you have black carpets or very dark flooring, the structured light sensor may be absorbed rather than reflected. Try placing a brightly colored marker near the toy to see if the robot recognizes the contrast; if it does, the issue is the floor's light absorption.
• Hard Reset the Sensors: Power down the unit completely using the physical power button, wait 60 seconds, and restart. This clears the temporary cache of the ReactiveAI processor and forces a re-calibration of the 3D sensors.

Heads Up

• Avoid using glass cleaners or alcohol-based wipes on the camera lens, as these can strip the hydrophobic coating and cause permanent streaking.
• The Q Revo MaxV struggles with transparent objects (like clear plastic bowls) and highly reflective surfaces (like chrome table legs), which can confuse the 3D depth perception.
• Regularly clear the side brushes of hair tangles; if the brushes are bogged down, the robot's movement becomes erratic, which can lead to the AI miscalculating the distance to an obstacle.
• Ensure the robot is not operating in "Do Not Disturb" mode during the day, as some users report that scheduled cleans during DND hours occasionally bypass certain high-energy AI processing tasks to reduce noise.