The batteries started to get a little tough towards 2019, so when we moved we took the opportunity to replace both NiMh packages with new replacement batteries. We now run it 4x / week, with about 10 minutes of maintenance below / month as part of our general cleaning routine and have still not had any major problems. Vacuum manually around the sensor and between the suction fan and filter.
Still no worries beyond what is stated below, maintenance now takes 5-10 minutes every month when you have learned (cleaning the brush drive belt, carefully around the sensor and wiping with a microfiber cloth). Runs 3-5 x / week. The machine has become next to indispensable, notices immediately when it has not been run in a day.
Still works really well, except that sometimes debris enters the brush drive system, resulting in "My brush is stuck". This consists of a motor hidden inside the robot + protruding pinion which in turn drives the brush with a strap. However, the entire system is sufficiently exposed to need to be cleaned (Takes about three minutes) with tweezers ~ every two months, depending on use.
Also clean quickly around the sensor with a soft f.d. camera brush, at right angles away from the sensor so that dust is not poked in.
All in all, cleaning and maintenance probably takes a maximum of 10-15 minutes every month. Considering that it runs ~ 15x 30 min during the same period, we consider it a very good investment. Noticeably less dust in the air, on surfaces, and inside the computer. ;)
The fix for the dust container seems to have held up without further ado. Have not noticed any problems with the battery (yet).
As pointed out, the sensor for the dust container is under criticism. As the dust container lacks a proper locking mechanism on the front (bumper side), it is only held in place by friction, which leads to the container coming loose with relatively small shocks in height.
- After a quick look at the construction, however, this could be remedied by locking the switch / microswitch (the protruding rubber / silicone plug between two small, silver-colored / unpainted screws) for the dust container in the closed position using a suitable piece of plastic. And with a few layers of silver tape on the outside of the dust container to give extra friction against the plastic wall, after this the problem disappeared completely!
But it is not _reasonable_ that such "ugly hacks" are needed! Minus that the error exists, and that the company has not come up with a solution, e.g. a self-locking container to remedy it.
Otherwise, the robot works really well. Interesting to follow it the first few times to see it navigate, which it does with an interesting combination of LIDAR and drive-in-things, at a fairly low speed. So far, it has not damaged anything. It also runs relatively slowly while it works, but can gain some speed when it gets to / from different places, as well as the charging station, which it also finds its way back to without further ado.
Note, however, that it is unavoidable to remove cables and things with low height (eg low bathroom scale) from the room, otherwise it risks climbing up and getting stuck.
Since the robot does not have access to _precisely_ everywhere, a traditional vacuuming may also be needed sometimes, but considerably less often, and when you do not have to vacuum the floor, it is much faster.
The cleaning ability is also surprisingly good, probably largely helped by the brush, but it is probably a good idea to quickly clean containers and filters after every / every other time to maintain suction power. When you get the "jerk" in, it only takes a few minutes.
Battery: Ni-MH, 14.4V, 3200-mAh (according to manufacturer) so 46 Wh, 12 cells, maybe Sub-C? Several on Amazon seem to complain that the batteries do not last long. Maybe it should be possible to replace it with four to five cells (13.2-16.5 V, 3500-2800 mAh) LiFePO4 + BMS? If it does not 'confuse' the robot too much. :)
"Hidden" USB and power port on the back, depending on when these supply power, maybe you could drive an action camera (Mobius with supercapacitor?), Mounted on top of the sensor, which is only activated when the vacuum cleaner is running?
Had hoped that someone fixed custom firmware you could have tinkered with, but apparently not ... yet? ;)