it keeps missing big spots in the middle of the room, just goes around them, the map looks fine.

Diagnosis

The Shark AI Ultra AV2511AE relies on a combination of a rotating LiDAR laser turret and infrared cliff sensors to map and navigate. When the robot "skips" large open areas despite the map appearing correct, it is experiencing "phantom obstacles." This occurs when the LiDAR laser reflects off a surface or is blocked by a smudge, leading the onboard processor to believe there is a physical barrier where none exists. Because the map is generated during the initial discovery phase, the robot "remembers" the room layout, but the real-time obstacle avoidance system overrides the map to prevent collisions. This is frequently caused by a buildup of fine dust or a "film" of oils on the LiDAR turret window or the front-facing proximity sensors. If the robot encounters a specific type of flooring—such as high-gloss dark tiles or mirrored furniture—the laser can be refracted or absorbed, creating a "blind spot" that the robot treats as a wall. You can confirm this by placing the robot manually in the center of the missed spot; if it immediately pivots and retreats without touching anything, the issue is definitely sensor-based rather than a mechanical wheel failure.

How to Fix It

To resolve phantom obstacles and restore full room coverage, follow these precise cleaning and calibration steps:

1. Power down the unit completely by pressing the power button and ensuring the LED indicators are off.
2. Locate the LiDAR turret on the top of the vacuum. Use a dry, high-density microfiber cloth to gently wipe the clear plastic perimeter of the turret window. Do not use glass cleaner or ammonia-based products, as these can streak the plastic and actually increase laser refraction.
3. Inspect the front bumper. Use a can of compressed air to blow out any debris lodged in the gap between the bumper and the chassis. If the bumper feels "stiff" or doesn't spring back instantly, gently tap around the edges to ensure the internal switches aren't stuck.
4. Clean the infrared cliff sensors located on the underside of the robot. Use a cotton swab dipped in a small amount of isopropyl alcohol (70%) to clear the dark plastic windows.
5. Perform a "Hard Reset" to clear the temporary obstacle cache. Turn the power switch off, wait 30 seconds, and turn it back on.
6. Clear the area of "invisible" obstacles. Remove any high-pile rugs with dark patterns or mirrored floor vases in the missed zone, as these are the primary triggers for the AV2511AE's avoidance system.

If That Didn't Work

• Rebuild the Floor Map: The existing map may contain a "ghost" wall saved from a previous run (e.g., a chair that was there once but is gone now). Go into the SharkClean app, delete the current map, and run a new "Map Only" cycle. This forces the LiDAR to re-index the room dimensions from scratch.
• Check for Lighting Interference: Direct sunlight hitting the floor at a low angle (especially during golden hour) can saturate the infrared sensors, making the robot think it is at a cliff's edge. Close the blinds or curtains in the affected area and run the vacuum again to see if the spots are covered.
• Inspect the Side Brushes: If the side brushes are tangled with long hair or string, the robot may struggle to "feel" the perimeter of the missed area, causing it to deviate from its path. Use a pair of scissors to clear the brush axles of all debris.
• Update Firmware: Check the SharkClean app for any pending firmware updates. Shark frequently releases navigation patches specifically to tune the sensitivity of the LiDAR avoidance algorithms to reduce false positives.

Heads Up

• Avoid using wax-based floor cleaners on your hard floors; the residue can create a mirror-like sheen that confuses the LiDAR laser.
• Never spray cleaning agents directly onto the LiDAR turret; always apply the cleaner to the cloth first to prevent liquid from seeping into the laser diode.
• If you have very dark charcoal or black carpeting, the robot may perceive it as a "cliff" (drop-off) and refuse to enter the area. This is a hardware limitation of infrared cliff sensors.
• Regularly check the LiDAR turret for "stuck" rotations. If you hear a clicking sound coming from the top of the unit, a foreign object may be obstructing the laser's rotation.