DJI Romo P2 Review: Excellent From Day One, DJI’s Robot Nears Perfection and Costs Less

The first DJI robot vacuum, the Romo, had impressed me with its navigation, far less with its price. With the Romo P2, the manufacturer refines its approach and returns with a substantially more reasonable price.

Fifteen days with a dog shedding, muddy footprints on the tiles, and dangling cables: the Romo P2 had plenty to handle. At the end of the test, it was, however, its way of moving among all these elements that impressed me the most. Approaching an object without pushing it aside, weaving between chair legs without hesitation, then resuming cleaning the stains: that seems obvious. After dozens of robots tested, I can assure you it isn’t obvious yet.

DJI again translates its drone know-how into cleaning our floors, with a launch price of €1,299, which is €600 less than the original Romo P. That’s enough to look at this second generation differently, which preserves its transparent design and motorized round mops.

I installed this premium version for fifteen days on roughly 45 m² of my ground floor, between tile and long-pile carpet. The opportunity to see whether DJI could still improve its first Romo, which had already set the bar quite high.

Design and construction: still transparent, a bit less rough around the edges

I really liked the look of the original Romo, and I remain as charmed by its successor. DJI has, however, revised the presentation of the visible components through the shell. The first one had a very bare-bones feel, almost an exposed assembly of parts, done to please the geeks. Which, I admit, worked quite well on me.

The P2 presents a more refined impression. The interior layout seems more carefully arranged, without losing what makes the device unique. You still see the guts of your vacuum, but the whole looks more elegant.

The footprint remains that of a premium maintenance station. DJI doesn’t magically remove the reservoirs and dust bag, but their use poses no problem. The tanks detach and fill easily, their handles are convenient, and I did not notice any drips during transport.

Under the robot, DJI preserves a dual-brush system for suction and two motorized mops for washing. The offset arm benefits from a stated 123° travel, with a reach increased compared to the first generation.

Suction: nothing scares it, not even the dog’s shedding

With a maximum suction power announced at 36,000 Pa, the Romo P2 carries solid specifications. But what matters most are the results on my tiled floor, and I have little to fault it for.

Dust, crumbs, and dog hairs are collected with impeccable efficiency. My dog chose the test period to molt, so the brushes had plenty to deal with.

The result is just as convincing along baseboards and in corners. The offset side brush reaches debris right up to the edges, leaving no small specks of hair behind that sometimes accumulate after a robot passes. In my usage, nothing remained on the tile.

The dual-brush system also handles tangling very well. After fifteen days, there were no mated tufts of hair to remove. The automatic emptying followed without difficulty, with no clog or waste buildup observed.

On long-pile carpet, the verdict remains very good, with the usual caveats. The robot automatically increases its power and takes time to work the surface. Cleaning is noticeable after its pass, but hairs and deeply embedded debris in the fibers do not all disappear. The Romo P2 remains among the most convincing models I’ve used on this terrain, though it cannot completely replace a handheld vacuum.

I mainly left it in automatic mode. The quiet mode is also satisfactory for daily maintenance, even if it seems less suited to a heavily soiled floor. To the ear, the robot remains quiet during suction.

Washing floors: the round pads still have plenty to offer

While rollers are gaining ground, the Romo P2’s two mops remind us that a classic solution can still work very well.

On fresh spills, the work is excellent. Oil, sauce on tile, footprints and autumn mud are removed during the automatic program. The robot usually re-passes over the stains, with about two passes observed, without me having to tell it to restart.

Dry stains help define its limits. Coffee is recovered well. A very ingrained sauce trace gives it more trouble: targeted extra cleaning may be necessary, and it remains sometimes easier to finish by hand. The P2 is a very good washer, but you cannot expect it to persist as long as a handheld washer would.

In the programs I tested, I found it slightly better than the Roborock Saros 20 Flow Complete, particularly on coffee. The DJI seems to better target certain traces and push more in automatic mode.

The second wash pass also contributes to this result: it picks up the remaining small traces and leaves a tile without haze or tacky sensation underfoot. Without this rinse, the marks are a bit more visible. I therefore prefer keeping it, even if it lengthens the cleaning. It’s really the entire program that convinces me more, without proving a superiority of the pads over the roller.

The offset mop cleans edges and around furniture legs properly. It does a good job, without dramatic differences compared to the best competitors. Finally, on the long-pile rug, the mop lifting proved sufficient: I did not feel moisture in the fibers after the robot passed.

There remains a small flaw, fairly common with obstacle recognition systems: a tuft of hair was bypassed instead of being sucked up. One instance during the test, but enough to remind that recognition isn’t infallible.

The lifting mechanism also allowed me to observe the robot traversing a small threshold, as well as crossing a chair leg, without particular difficulty. Since the height hasn’t been measured, this doesn’t validate the limits DJI announced.

The robot pairing works without a hitch. DJI Home provides a visual interface, tutorials properly explain the functions, and the translation has improved since my first test.

However, program management remains the weak point of the experience.

To choose the wash type, to select the cleaning solutions, or to initiate the second rinse, you always have to enter the program settings. The possibilities are numerous, but their organization still isn’t intuitive. Even in automatic mode, I find myself fiddling with the settings to ensure I get what I want.

Once the program is configured, the robot works very well. I would have simply liked DJI to make this setup more obvious. In this regard, Roborock still offers a more satisfying experience in my view.

Reliability isn’t in question. I encountered no bugs, no disconnections, and no cleaning interruptions during these fifteen days. The scheduling I experimented with executed without issue. The remote camera was used, unlike the voice commands, which I did not test.

DJI also offers a Local Data mode, which allows continuing to use the Romo P2 without an Internet connection after initial configuration. Proximity control then goes through the phone’s Bluetooth. According to the manufacturer, maps, images, and navigation data remain stored and processed in the robot rather than in the cloud. A welcome option for those who prefer not to entrust their home plan to a distant server than to let a vacuum pick up their crumbs.

With the program that includes the rinse with clean water, the Romo P2 takes about 1 hour and 20 minutes to cover my 45 m². It’s quite long for a space of that size.

I counted nearly five returns to the station to wash the mops during the cycle. Skipping the second rinse saves about half an hour, at the trade-off of more visible traces described above. So duration remains a consideration if you want to start the robot between moments of life in the home.

The station requires a fresh water refill every two to three washes in my experience. Its automatic maintenance seems more effective than that of the first Romo: the tray and the mops generally stay clean, and I didn’t notice any odor.

Nevertheless, there are still some hairs and small debris after the cycles. Self-cleaning does not eliminate the need to inspect the pads and periodically clean them yourself, just like the washing tray. These fifteen days provide a solid first impression, but don’t allow judging wear over several months.

Drying takes about three hours. The blower is audible, but it remains unobtrusive if the station is placed far enough from the living room. At the end of the program, the mops are perfectly dry.

DJI Romo P2: Clubic’s verdict