The AI piloting the X11 occasionally showed some strange behavior. A couple of times, the robot cleaned my whole place, then went back to one random spot before turning around, heading back to its base, and calling the job complete. Bear in mind, it didn’t actually vacuum that extra spot. The robot seemed to just need a little extra exercise.
The X11 also sometimes stopped for a while to charge, then cleaned one more small area before the job was complete. It clearly had enough battery to just finish the job in the first place.
For measured testing, I close the robot in two rooms of my apartment, one with carpet and the other with hardwood. I then sprinkle a specific amount of rice (representing everyday debris) and then sand (representing stubborn debris) on the floor, and weigh the dustbin before and after each cleaning run to see how much the robot picked up.
After moving the unit for these tests, the X11 was unable to find its base station in these new rooms, no matter how many times I picked it up and pointed it in the right direction. It kept trying to escape back to the kitchen, where I initially set up the base. I ended up needing to make new maps for each individual room, and the robot even once switched back to the old map in the middle of a run and tried to escape to the kitchen again. Given my unusual testing procedure, I don’t think this last issue will come up very often in everyday circumstances, but it is telling of the robot’s sometimes wonky AI.
That same AI also caused it to underperform when picking up sand on carpet, a suction power stress test. The robot ran over a cord at one point in the testing process, a mistake on my part, but when I reran the test, it kept increasing and then lowering its suction power as though it was hesitant to go all out.
I ran the test a few times to see if it would make an adjustment. It eventually smoothed out the suction, but the AI never detected the extra dirt and turned up the power accordingly. As a result, the X11 did worse on my suction stress test than both of its predecessors. It picked up just 23.35% of sand on carpet compared with 37.7% for the X8 and 41.68% for the X9.
The X11 is one of the rare vacuums to score a perfect 100% when picking up rice on carpet, but the X8 and X9 achieved better cumulative cleaning scores across both surfaces. The X11 fared well on hard flooring, picking up 91.35% of rice and 66.25% of sand. It didn’t seem to fling any particles on hard flooring, a common behavior in other high-end robot vacuums.
The X11’s cleaning pace is similar to its predecessors. It finished all carpet tests in an average of 11 minutes and 27 seconds and hardwood tests in 14 minutes and 29 seconds. I didn’t notice any missed spots either, so it’s a thorough and efficient navigator, even if the AI can sometimes confuse it.
During my mopping stress test, in which I spread a quarter-ounce of raspberry jelly on my kitchen floor, the X11 redeemed itself with an excellent result. It cleaned up the entire stain, leaving only a tiny smudge and no seeds, residue, or stickiness spread to other parts of the floor. The base station also thoroughly cleaned the mop after the run, leaving no visible remnants of the difficult test.
Both the X8 and the X9 fared well on this test, too, but the latter left two stray seeds, a noticeable splotch, and a touch of sticky residue in the surrounding areas. The X8 missed three seeds, but left no residue. The X9 washed away all evidence of the test when back at its base, like the X11, while the X8 still had a few seeds left along the spindle.
