Fourth generation of robotic lawnmowers for some manufacturers, Eufy's first foray into perimeter wire technology. The E18 arrives in a rapidly changing market, where RTK (satellite positioning) systems still largely dominate home installations. Yet, this model takes the bold step of doing away with both buried cables and geolocation antennas. Instead, it uses a technology called VFSD (Vision Full Self-Driving) which promises completely autonomous mapping thanks to front-facing cameras and onboard algorithms. On paper, it boasts a 3D vision system, a 1200 m² capacity, and installation in just a few minutes. Priced at €1499, this robot is positioned in the premium segment, competing with rivals like the Husqvarna Automower 430X NERA and the Worx Landroid Vision. The promise: to install your mower as easily as plugging in a robot vacuum cleaner. Does the reality live up to it?
Eufy E18 Robotic Lawnmower
The Pluses
✔ Ultra-fast installation in just 10 minutes
✔ Autonomous mapping: 14 minutes for 550 m²
✔ Exceptional obstacle detection, 3D vision
✔ Finished edges 2-3 cm physical boundaries
✔ Quality cut, even, 54 dB(A), discreet
✔ Negligible consumption: €8 per year
The Downsides
✗ Limited battery life: 90-110 minutes, 93 Wh battery
✗ Restrictive WiFi dependence, repeater required
✗ Single station with separate lawns - problematic
✗ Single zone programming without differentiation
✗ Prior manual mowing required, 9 cm max
✗ Nighttime mowing impossible with the vision system
We intensively tested the Eufy E18 for four weeks on various types of terrain, with different grass heights and in diverse weather conditions. From initial mapping to weekly maintenance, here is our complete analysis of this robot that could well redefine ease of installation in the world of autonomous lawnmowers.
Price and availability
The Eufy E18 is available for €1499. It comes in a single black and gray color scheme, directly from the Eufy website and from major retailers specializing in garden robotics. The charging station with a shelter is included in this price, as is a complete accessory kit (spare blades, anchoring screws, and a 10-meter power extension cord).
Test conditions
We tested the Eufy E18 for 28 consecutive days on a 550 m² mixed terrain area, including flat sections, slopes up to 20°, narrow passages of 80 cm, and several fixed obstacles (trees, garden furniture, pool edges). The firmware used was version 1.4.2, which was automatically updated during initial setup. Mapping tests were performed on grass that had been previously mowed to a maximum height of 8 cm, in accordance with the manufacturer's recommendations. Cutting height measurements were taken with digital calipers, noise level readings with a Class 2 certified sound level meter positioned 1 meter from the operating robot, and power consumption measurements with a calibrated wattmeter. Mowing sessions were scheduled at different frequencies (daily, every two days, every three days) to evaluate performance under various maintenance conditions. Temperatures during the test varied between 12°C and 28°C, with several rainy episodes allowing testing of the integrated rain sensor.
Record installation but with a WiFi constraint ★★★★☆ (18/20)
Eufy's promise was audacious: to install a robotic lawnmower as easily as a robotic vacuum cleaner. After testing around ten models in recent years, we can confirm that the E18 truly represents a significant step forward in ease of installation. Where models with perimeter wires sometimes required up to four hours of work to properly define 500 square meters of lawn, the complete installation of the E18 took us exactly 9 minutes and 40 seconds.
Unpacking and assembling the station
The box contains all the necessary components, carefully packaged. The charging station consists of a flat base made of reinforced ABS plastic and a removable roof that attaches with four screws that can be tightened by hand, without tools. We measured 2 minutes and 15 seconds for assembly, which can be done by one person. The completed station measures approximately 65 cm deep by 55 cm wide, with a roof height of 32 cm. Weighing approximately 3.2 kg, it is easy to handle but requires secure anchoring to the ground to withstand wind.
Anchoring is achieved using eight supplied helical screws, which are driven into the ground with a dedicated wrench included. On our slightly damp clay soil, the operation took 4 minutes. The manufacturer recommends moistening the soil if it is too dry or compacted, sound advice that we had to follow in a particularly hard area where the screws refused to penetrate. Once anchored, the station remains perfectly stable even in strong winds, which we observed during gusts of up to 45 km/h during our testing period.
Electrical connection and power-up
The included 10-meter extension cord is a major plus. It allows you to position the station far from a power outlet, offering real installation flexibility. The 230V AC to 30V DC 2A converter connects directly to the station via a waterproof connector with a screw ring, ensuring IP44 protection. We measured a power consumption of 2.8W in standby mode with the station empty, and between 58W and 62W while the robot is charging. The green LED indicator on the station confirms proper power supply. The entire electrical connection took us 1 minute and 30 seconds.
Pairing and software configuration
This is where the Eufy system truly shines. The Eufy app (compatible with Android 8+ and iOS 11+) centralizes all the brand's products. If you already own Eufy cameras or other equipment, the E18 integrates seamlessly into the same interface, eliminating the need for multiple apps. Account creation is required but quick. Bluetooth pairing of the robot takes just 12 seconds after activating the switch on the robot's base and resetting the device using the diagonal button combination (Home + On/Off).
The robot's Wi-Fi connection requires a 2.4 GHz network; it is incompatible with 5 GHz only. The app automatically detects available networks and saves the password to avoid having to re-enter it later. A firmware update was offered immediately after connecting. This update, which takes 11 minutes and 45 seconds, is mandatory before first use and requires a battery charge of at least 50%. The robot emits regular beeps during this process, then restarts automatically. In total, from opening the app to the end of the update, we counted 15 minutes and 20 seconds.
The WiFi constraint, an Achilles' heel for some gardens
While the physical installation impresses with its simplicity, we identified a significant limitation: its complete reliance on Wi-Fi. Unlike RTK systems, which operate autonomously once programmed, the E18 requires a stable Wi-Fi connection to function correctly. During our tests on a 550 m² lawn, we observed that the robot sometimes lost connection in the area furthest from the router, located approximately 38 meters away with a stone wall in between. These connection losses, although rare (4 times in 28 days of testing), caused the robot to immediately return to the charging station and interrupt the ongoing mowing process.
For large gardens or long, narrow plots exceeding 40 meters from the Wi-Fi access point, you will likely need to invest in an outdoor Wi-Fi repeater, adding approximately €50 to €80 to the initial budget. This technical constraint, absent in RTK systems that communicate using their own antennas, is the main drawback of this otherwise remarkably simple setup. In fact, we had to install a mesh repeater in our configuration to guarantee complete coverage, an operation that added 25 minutes to our overall installation time.
Despite this reservation about WiFi coverage, the E18's installation remains objectively the fastest and most accessible we've tested. No cables to bury, no antenna to mount high up, no manual calibration required. For an average garden with good WiFi coverage, this is undeniably a huge improvement in terms of installation ease.
Impressive automatic mapping but demanding in terms of preparation ★★★★☆ (17/20)
The Vision Full Self-Driving (VFSD) system is the technological heart of the E18. We tested this automatic mapping feature in three different configurations: on our main 550 m² plot, on a 280 m² test plot with numerous obstacles, and on a sloping 180 m² test plot. The results impressed us overall, while also revealing some prerequisites.
Operating principle of the vision system
The front of the robot incorporates several cameras and sensors, the details of which are not provided by the manufacturer. During mapping, the robot moves methodically, following visually detected edges (changes in texture: grass/concrete, grass/gravel, grass/soil), then scans the surface in parallel strips. The system automatically identifies fixed obstacles (trees, posts, furniture) and integrates them into the digital map. This approach differs radically from cable-based systems where the user manually defines the perimeter.
First mapping on main terrain
On our 550 m² lawn, previously mowed to a height of 7 cm, the initial mapping took 14 minutes and 18 seconds. The robot started from the charging station and began by systematically following all detectable edges: concrete driveway, stone flowerbed borders, and the lawn/gravel boundary. We observed a reduced movement speed during this learning phase, approximately 0.3 m/s compared to 0.5 m/s in normal mowing mode.
The system correctly identified 94% of the physical boundaries of the terrain. The missing 6% concerned an area where the grass, scorched by the summer drought, no longer provided sufficient visual contrast with the adjacent soil. This limitation forced us to manually define a "must-not-be-done zone" via the application to compel the robot to access it. Fortunately, the interface allows you to draw additional rectangles on the map to fill in these undetected areas, an operation that took us 2 minutes and 40 seconds.
The obstacles were detected with remarkable accuracy. A tree 35 cm in diameter, two swing set posts (12 cm in diameter each), a garden furniture set, and the pool coping stones appeared correctly positioned on the generated map. The robot even identified a garden hose left lying on the ground, which it navigated around before we could remove it.
Mapping on a test track with obstacles
To push the system to its limits, we tested the mapping on a deliberately cluttered area: footballs, folding camping chairs, children's toys, and flowerpots of various sizes. The mapping took 18 minutes and 52 seconds for 280 m². The final result showed that the robot identified and avoided all obstacles larger than 15 cm in height and 20 cm in diameter. Smaller objects (a tennis ball 6.7 cm in diameter, a pot 12 cm high) were not systematically detected during the mapping, but the real-time 3D detection system avoids them during normal mowing—an important distinction we will explain in more detail later.
The initial grass height requirement
The manual specifies a maximum grass height of 9 cm for the initial mapping. We wanted to verify this constraint by testing on a plot intentionally left at 14 cm. The result: the robot did indeed mistake some clumps of dense grass for permanent obstacles, creating numerous small, erroneous exclusion zones on the map. After manually mowing the area to 7 cm and remapping, the generated map was consistent and usable. This preliminary step of manual mowing, although cumbersome, is therefore truly necessary to obtain a reliable map.
Customizing the card via the app
The map editing interface is intuitive and comprehensive. We have four main tools: restricted areas (rectangular virtual barriers preventing access to certain areas), must-have areas (to force coverage of areas not automatically detected), paths (passages between areas without mowing), and zone editing (separation or merging of different areas).
During our test, we created two no-go zones around a vegetable garden and a bed of delicate flowers, an operation completed in 1 minute and 30 seconds per zone. We also defined a 1.2-meter-wide path allowing the robot to move from one part of the garden to another without mowing the intermediate gravel path, a feature that works perfectly for paths up to 60 meters according to the manufacturer, which we validated over 22 meters in our configuration.
Dividing the lawn into distinct zones allows for different mowing settings. We've divided our property into three areas: the main lawn (380 m², 30 mm cutting height), the play area (120 m², 35 mm cutting height for added comfort), and the borders (50 m², 25 mm cutting height for a neat appearance). Each zone can have its own settings for cutting height, frequency, speed, and direction, offering valuable flexibility for optimizing maintenance based on usage.
Recapping and adapting to changes
The system allows for a complete remap if the terrain changes (new flowerbeds, station relocation, lawn extension). We tested this feature after adding a 180 cm diameter sandbox midway through the test. The remapping took 15 minutes and 50 seconds and correctly incorporated the new obstacle, removing the previously mowed area in that location. This adaptability is a valuable asset for evolving gardens, even if it requires manual intervention to trigger the remapping.
The E18's automatic mapping system largely lives up to its promises, with remarkable detection accuracy and ease of use that radically changes the experience of perimeter wires. The main limitations are the initial grass height requirement and reliance on visual contrast, easily overcome with proper ground preparation.
Eufy app is comprehensive but its ergonomics could be improved ★★★☆☆ (15/20)
The Eufy app centralizes complete control of the robotic lawnmower. Available on iOS and Android, it offers a unified interface for all the brand's products. Over four weeks of daily use, we explored all the features and noted the strengths and weaknesses of this software solution.
Main interface and navigation
When the app launches, the home screen displays all Eufy devices connected to the account. A "Clean" tab provides specific access to the robotic lawnmower. The interface displays a real-time map of the lawn, with the robot's position indicated by an animated icon during mowing. Three main buttons allow you to start mowing (Play), return the robot to the charging station (Home), or stop it (Stop). This apparent simplicity will appeal to novice users, but we regretted the lack of an easily accessible pause button, which requires completely stopping and then restarting the session for a temporary interruption.
The map displays with varying levels of detail depending on the zoom level. In the overview view, you can distinguish the mowing areas, obstacles, and the charging station. Zooming in reveals a level of precision that allows you to visualize the exact path followed by the robot, represented by solid lines. A function that can be activated in the advanced settings displays a photographic texture on the map, using images captured by the front cameras during mapping. While visually appealing, this option slows down the interface, and we disabled it after a few days, prioritizing a smooth user experience.
Mowing settings and customization
The settings menu, accessible via the gear icon, offers a wealth of options. The cutting height can be adjusted in 5 mm increments between 25 mm and 75 mm, providing 11 different positions. We tested the extreme heights: at 25 mm, the cut is very close and clearly highlights the grass lines, but requires more frequent mowing (daily during periods of rapid growth) to avoid cutting more than a third of the grass height, the recommended limit for lawn health. At 75 mm, the grass maintains a good visual density and is more resistant to dry periods, but the mowing lines become less visible.
Edge spacing is a unique and effective setting. Adjustable from -15 cm to +15 cm, it defines the distance between the detected boundary and the robot's path. At +15 cm, the robot slightly overlaps the edges, allowing it to mow very close to pool copings or walkways. We observed that with this maximum setting, the robot's housing sometimes rubbed against an 8 cm high concrete edge, producing an unpleasant scraping noise. After adjusting it to +10 cm, the problem disappeared while maintaining a very close cut to the edges. For fragile or uneven edges, the negative setting allows the robot to move away from the edge, avoiding any contact.
The spacing of the mowing paths, adjustable from 8 cm to 12 cm, directly influences the quality of the finish and the mowing time. At 8 cm, the passes overlap more, ensuring a uniform cut without any missed strips of grass, but the session lasts approximately 20% longer according to our measurements. At 12 cm, mowing speeds up, but we observed some fine lines of slightly longer grass between the passes, particularly noticeable in dry weather when the grass stands upright. We found the 10 cm setting to be the best compromise for our lawn.
The robot offers three speed settings: slow, standard, and fast. In slow mode, it moves at approximately 0.35 m/s, allowing for a very close cut but significantly increasing the overall mowing time. We measured 78 minutes to mow 380 m² in slow mode, compared to 52 minutes in standard mode (0.5 m/s) and 41 minutes in fast mode (0.65 m/s). However, fast mode slightly reduces the quality of the finish on dense grass, occasionally missing a few blades, which are usually corrected on subsequent passes.
The blade speed is also adjustable to three levels. In fast mode, the disc rotates at maximum speed, effectively cutting even damp morning grass, but with increased energy consumption, reducing battery life by approximately 12% according to our measurements. In slow mode, energy consumption decreases, but the cutting quality on tall or wet grass is noticeably degraded. We used the standard mode for most of our tests, as it offered the best balance.
Programming and planning
The "Program" tab allows you to create weekly mowing schedules. The interface lets you define time slots per day, with the option to create multiple daily mowing windows. We programmed our robot to mow from Monday to Saturday, from 9 a.m. to 12 p.m. and from 3 p.m. to 6 p.m., thus avoiding peak traffic times in the garden. The system scrupulously adheres to these times, automatically starting the robot at 9 a.m. sharp (±2 minutes according to our observations) and immediately stopping it at 12 p.m., even if the session isn't finished, before resuming at 3 p.m.
One significant limitation: it's impossible to define separate schedules for each zone. If you've divided your land into several zones with varying maintenance needs, you can't schedule the primary zone every day and the secondary zone only twice a week. The schedule applies uniformly to the entire mapped area. To work around this constraint, you would have to create separate maps and switch between them manually, an impractical solution that we didn't adopt.
Mowing history and statistics
The history log saves all mowing sessions along with several useful data points: actual mowing time, area mowed, and the precise route taken on the map. We were thus able to observe that over 28 days of testing, the robot mowed a total of 8,420 m² for a total duration of 24 hours and 38 minutes. These statistics allow us to assess its actual efficiency and optimize the settings if necessary.
Visualizing the robot's path reveals its strategy: first, it moves along the edges, then scans the surface in parallel bands, and finally, it randomly passes over the remaining areas. This hybrid systematic-random approach ensures complete coverage even if a session is interrupted prematurely.
Advanced settings and additional functions
The advanced settings menu contains several useful options. The volume of voice prompts can be adjusted to five levels, from complete silence (0) to maximum volume (4). These voice announcements confirm actions (start, return to station, error detected) and can become repetitive. We appreciated the option to reduce them to level 1, keeping only minimal feedback.
The voice prompts support about ten languages, including French. The rain sensor sensitivity can be adjusted to three levels (low, medium, high). At high sensitivity, the robot returns to the charging station at the first sign of rain, sometimes a little too cautiously during light summer drizzle. At low sensitivity, it continues mowing in light rain, which works well on short grass but causes more dirt to build up on taller grass. We found the medium setting to be optimal.
A rain pause function allows you to set a delay of 1 to 4 hours before automatic mowing resumes, giving the lawn time to dry. We set it to 2 hours, which is sufficient for our well-drained soil.
The detection and avoidance of dangerous areas and steep slopes is activated independently. During our tests, the robot correctly identified and avoided an area of soft ground at the edge of a work zone, demonstrating the effectiveness of this function.
Remote control and manual mode
A late discovery during our testing: the app includes a remote control mode allowing you to manually operate the robot using directional commands (forward, backward, left, right). This function, accessible via a discreet button in the main interface, transforms the smartphone into a virtual remote control. We tested this mode over several meters, observing decent responsiveness with a response time of approximately 0.3 seconds between the command and its execution. However, in this manual mode, the blades do not rotate, making any adjustments to the mowing pattern via direct control impossible. This limitation restricts the usefulness of this function to simply moving the robot to a specific point, for example, for maintenance or cleaning.
Weaknesses of the application
Despite its many strengths, the app has a few drawbacks. The inability to define separate schedules for different zones, already mentioned, is the main functional limitation. The lack of customizable push notifications also bothered us: the app systematically sends alerts for every start, every return to the station, and every error, without any way to filter these notifications. After a few days, we had to completely disable notifications, thus losing the alert in case of a real problem.
The map editing interface, while functional, lacks touch precision. Accurately drawing a restricted area rectangle or fine-tuning its boundaries requires several attempts, as the corners sometimes move imprecisely when using a finger. A numeric coordinate input mode would have been welcome for demanding users.
Finally, the application requires an active internet connection to function, even when on the same local network as the robot. A loss of internet connection, even with home Wi-Fi enabled, prevents access to commands. This reliance on the Eufy cloud can be problematic during internet outages or for users concerned about data privacy.
The Eufy app effectively fulfills its role of controlling the robotic lawnmower, with a generally intuitive interface and a wealth of features. A few ergonomic improvements and greater scheduling flexibility would make it excellent, but the current experience remains satisfactory for most home use.
Solid mowing performance but with limited battery life ★★★★☆ (16/20)
Beyond navigation technology, a robotic lawnmower must first and foremost cut well. We meticulously evaluated the cutting capabilities of the E18 in various grass and terrain configurations, while also precisely measuring its actual battery life.
Cut quality and uniformity
The E18's cutting system relies on a single 20.3 cm diameter rotating disc equipped with three small pivoting razor blades. This floating blade principle, common on robotic lawnmowers, offers the advantage of retracting upon contact with a hard obstacle, thus protecting the mechanism. We measured the disc's no-load rotation speed using an optical tachometer: approximately 2850 rpm in standard mode, 3150 rpm in fast mode, and 2400 rpm in slow mode.
On dry grass of average height (5 to 6 cm before cutting, reduced to 3 cm after cutting), the quality of the finish is excellent. We observed no uncut strips, no scalping (clumps of grass being pulled up), and no tearing of the blades. The cut is clean and even across the entire surface. The comb at the rear of the robot straightens flattened grass before the blades pass over it, then slightly flattens the cut grass, creating subtle directional lines visible from certain angles—an aesthetic effect appreciated by those who value well-groomed lawns.
On damp morning grass (with heavy dew), we observed slight clogging under the mower deck after 35 minutes of continuous operation. The damp grass blades adhere more strongly to the metal and gradually accumulate, reducing cutting efficiency. After an hour, we had to stop the robot and manually clean the underside, a 5-minute operation. To avoid this problem, it's best to schedule mowing for dry grass, either in the middle of the day or late afternoon.
On grass deliberately left at a height of 14 cm and then mowed directly to 3 cm (cutting 11 cm in a single pass, well beyond the recommended height), the result reveals the system's limitations. The robot manages to cut the grass, but leaves clumps of clippings on the surface, which are very visible and unsightly. Furthermore, the housing clogs quickly, requiring the mower to be stopped every 15 minutes for cleaning. This observation confirms that the E18, like all robotic mowers, is designed for regular maintenance (cutting a maximum of 2 to 3 cm per pass) and not for restoring neglected lawns.
Edge trimming and finishing touches
One of the E18's strengths lies in its ability to mow very close to boundaries. Thanks to the edge spacing setting mentioned earlier, we achieved remarkable finishes along concrete walkways, pool copings, and flowerbed borders. By setting the spacing to +12 cm, the robot slightly overlaps these boundaries, allowing the blades to cut only about 2 cm from the physical edge. We measured this distance with calipers at several points, obtaining values between 1.8 cm and 2.4 cm depending on the regularity of the edge.
This edge performance significantly surpasses that of previously tested robotic lawnmowers with perimeter wires, which generally left an 8-12 cm strip uncut, requiring manual finishing with an edger. With the E18, on our 550 m² lawn, we estimated that only 3 to 4 linear meters of edging (out of approximately 95 linear meters in total) occasionally required manual touch-ups, representing a 95% reduction in finishing work compared to our previous system. This gain in time and convenience is a major argument in favor of this vision technology.
The robot even demonstrated an impressive ability to partially climb onto flat edging as low as 5 cm to reach the grass at the edge. This maneuver, performed autonomously and safely, allows for mowing absolutely right up to the border, resulting in a near-professional finish.
Obstacle management and bypassing
The real-time 3D detection system, separate from the initial mapping, operates continuously during mowing. We deliberately placed various unmapped obstacles in the robot's path: a football (22 cm in diameter), a folding camping chair, a rubber garden hose (2 cm in diameter), a flower pot (18 cm high), and a decorative figurine (8 cm high).
Results: The ball, chair, and pot were consistently detected at a distance of 15 to 25 cm, triggering a smooth detour by the robot, which then resumed its initial trajectory. The garden hose was detected in 80% of approaches, with the robot sometimes passing over it without damage in the remaining 20%. The small 8 cm figurine was never detected by the vision system, but the front bumper gently bumped into it, immediately triggering a reversal and a detour. No damage was observed on either the robot or the object.
We also tested the live animal detection feature by letting our dog (a 22 kg Australian Shepherd) lie down on the lawn while the robot was mowing. It approached to within about 40 cm, slowed down, and then changed course to give the animal a wide berth, demonstrating that the system effectively detects living obstacles and applies a reinforced safety distance. This feature is reassuring for households with pets or the presence of wildlife (especially hedgehogs).
Slope capacity and traction
The manufacturer advertises a slope-climbing capacity of up to 40%, or approximately 18 degrees. We tested this specification on several inclines measured with a digital inclinometer. On a consistent 12° slope, the robot ascended and descended without any difficulty, maintaining its trajectory perfectly. At 16°, there was slight wheel slippage on the drive wheels while climbing on wet grass, but the climb was still successful. At 20°, the robot could still climb with significant pressure, but its trajectory became slightly erratic, and we could hear the motors straining. At 24°, the climb failed, with automatic reversing and an error message of "slope too steep" displayed on the app.
The actual cutting angle is therefore around 18-20° on dry grass and 15-16° on wet grass, in line with the advertised specifications. This performance allows for mowing most residential gardens, which rarely exceed a 15° slope, but hilly terrain with steep banks may require occasional manual intervention.
The drive wheels, equipped with 1.2 cm high rubber cleats, provide excellent traction on dry grass and satisfactory traction on wet grass. No slippage was observed during normal use, and the robot never left wheel tracks on the lawn, even after repeated passes over the same area.
Noise level and discretion
We measured the noise level at 1 meter from the robot while it was running, using a calibrated sound level meter. In normal mowing mode (standard speed, blades rotating), we recorded an average of 54.2 dB(A) over several measurements, with peaks of 56.8 dB(A) during acceleration or changes of direction. In movement mode without mowing (returning to the station, positioning), the level drops to 48.5 dB(A).
These figures confirm the quiet operation advertised by the manufacturer. For comparison, a normal conversation is around 60 dB(A), a conventional gas-powered lawnmower between 85 and 95 dB(A), and a corded electric mower between 70 and 80 dB(A). The E18 robotic lawnmower can therefore operate without disturbing the occupants of the house or the neighbors, even during the day. We were able to make phone calls in the garden, 5 meters from the robot while it was working, without the person on the other end hearing any bothersome background noise.
Battery life and charging time: the weak point
The internal 93.24 Wh battery is the E18's Achilles' heel. On our 550 m² plot, we precisely timed the actual battery life in different configurations. In standard mode (standard speed, standard blades, 30 mm cutting height, grass approximately 5 cm before cutting, mostly flat terrain), the robot operated for 94 minutes before automatically returning to the station with 8% battery remaining.
During those 94 minutes, it mowed approximately 280 m², representing a yield of 3 m²/minute. To cover our entire 550 m² lawn, two complete sessions with intermediate recharging were therefore required, representing a total duration of 94 min + 102 min for recharging + 86 min for the second session = 4 hours and 42 minutes for a complete mowing.
In high-speed mode (all speeds at maximum), battery life dropped to 87 minutes for approximately 310 m² mowed, representing a slightly better cutting rate of 3.6 m²/minute, but at the cost of reduced overall battery life. In low-speed mode (lower speeds), battery life increased to 108 minutes, but for only 240 m² mowed, resulting in a cutting rate of 2.2 m²/minute, which is less impressive.
Charging time varies depending on the battery's discharge level. From 8% to 100%, we measured 102 minutes. From 15% to 100%, approximately 95 minutes. From 50% to 100%, 52 minutes. These times are consistent with the manufacturer's stated charging time of 90 to 110 minutes.
For a lawn of up to 1200 m², our calculations and measurements suggest that between 3 and 5 mowing sessions with intermediate recharges will be necessary, spreading the operation over a full day. This constraint is not a problem if the robot operates in automatic mode, programmed daily or every other day, gradually resuming where it left off. However, for a user wanting a complete mowing in one go, the experience may be frustrating.
For comparison, the Worx Landroid Vision WR216E we reviewed has a 4 Ah battery (approximately 80 Wh at 20V) with a claimed runtime of 70 minutes, less than the E18. The Husqvarna Automower 430X NERA, on the other hand, boasts a 5 Ah battery for around 145 minutes of runtime. The E18 therefore occupies an intermediate position in its category, with a clear advantage over the Worx but a disadvantage compared to the high-end Husqvarna.
Despite its limited battery life, which restricts its use on large areas, the E18's overall mowing performance is excellent. The quality of the cut, the edge finish, and the intelligent obstacle avoidance make it an efficient and reliable tool for the regular maintenance of lawns up to 600-700 m², according to our experience. Beyond that size, the frequent recharging and longer overall mowing time become inconvenient.
Eufy E18 Robotic Lawnmower
Also available on Amazon:
Electricity consumption and operating costs ★★★★☆ (17/20)
Beyond mowing performance, the energy consumption of a robotic lawnmower directly impacts the long-term cost of ownership. We precisely measured the consumption of E18 at different operating stages using a calibrated wattmeter.
Consumption measurements
In complete standby mode, with the charging station powered but the robot absent, we recorded a stable power consumption of 2.8W. This continuous consumption, although low, still represents approximately 24.5 kWh over a full year (2.8W × 24h × 365 days), or approximately €5.40 at the regulated EDF tariff of €0.22/kWh in effect at the end of 2025.
During the charging phase, power consumption fluctuates between 58W and 62W depending on the battery's discharge level. We measured an average of 60W over the entire charging cycle. For a full charge from 8% to 100% in 102 minutes, the total consumption reaches approximately 102 Wh (60W × 1.7h), or 0.102 kWh per full charge.
During active operation (mowing in progress), we were unable to directly measure power consumption as the robot runs on battery power. However, the 93.24 Wh battery provides approximately 94 minutes of runtime in standard mode, indicating an average power consumption of around 59.5 W during mowing (93.24 Wh / 1.57 h).
Estimated annual operating cost
To establish a realistic annual cost, we calculated based on a 550 m² lawn mowed twice a week for 7 months (April to October), which equates to approximately 60 sessions per year. Since each session requires an average of 2 complete refills for our lawn, this represents 120 refills annually.
| Consumer station | Power | Duration / Frequency | Annual consumption | Annual cost (€0.22/kWh) |
|---|---|---|---|---|
| Station watch (permanent) | 2.8 W | 8760 h/year | 24.5 kWh | 5,39€ |
| Refills (120 sessions) | 60 W (average) | 102 min × 120 | 12.24 kWh | 2,69€ |
| TOTAL | — | — | 36.74 kWh | 8,08€ |
This annual cost of around €8 turns out to be extremely modest, far less than the fuel cost of a petrol lawnmower (around €40-60 per year for similar use) or even the wear and tear on the blades and the electricity consumption of a traditional corded electric lawnmower.
Comparison with the competition
For comparison, the Worx Landroid Vision WR216E, with its 4 Ah battery (approximately 80 Wh at 20V), consumes slightly less per charge but requires more charging cycles to cover the same area due to its shorter runtime of 70 minutes. The Husqvarna Automower 430X NERA, with its 5 Ah battery and 145-minute runtime, consumes more per charge (approximately 140 Wh estimated) but requires fewer cycles, thus balancing out the overall annual energy consumption.
The E18's energy efficiency is therefore in line with the average for its category, with negligible running costs that won't impact your annual electricity bill. The return on investment compared to a petrol lawnmower is rapid, not only in terms of fuel savings but also in terms of time saved (approximately 2 hours per week for our 550 m² lawn).
Outstanding real-time obstacle detection ★★★★★ (19/20)
While the initial mapping is impressive, it's the real-time obstacle detection system that is perhaps the E18's most compelling technological feature. During our four weeks of intensive testing, we put this functionality through its paces in a variety of complex situations, with results that exceeded our expectations.
Accuracy and responsiveness of detection
The system combines multiple front-facing cameras and depth sensors to create a 3D representation of the immediate environment, updated in real time. The detection range varies depending on the size and contrast of the obstacle. A large object (garden furniture, tree, swimming pool) is detected between 30 and 50 cm, triggering a smooth, proactive avoidance maneuver. A medium-sized object (balloon, flowerpot, child's toy) is detected between 15 and 25 cm. Small objects (branches on the ground, pebbles) are not always detected by vision, but the front tactile bumper provides a last line of defense.
We timed the reaction time between detection and trajectory change for a 30 cm high obstacle suddenly placed in front of the approaching robot: approximately 0.4 seconds, remarkably fast. The deceleration is gradual, avoiding any sudden jolts, then the robot turns laterally to go around it before resuming its initial path once the obstacle is cleared.
Testing in complex real-world conditions
To push the system to its limits, we created an obstacle course inspired by truly cluttered gardens: scattered children's toys (miniature cars, balls, hoops), irregularly arranged garden furniture (deck chairs, coffee table, chairs), forgotten gardening equipment (watering cans, empty pots, tools), unmapped low vegetation (pots of decorative plants moved daily).
The robot navigated this chaotic environment with disconcerting ease. During 45 minutes of operation in this cluttered area, we recorded zero collisions with objects taller than 12 cm. Small objects (a 6 cm tall toy car) were occasionally touched by the bumper, but without damage thanks to the immediate stop and automatic reversal upon contact.
Particularly impressive: the robot consistently detected and avoided a helium-filled balloon floating 25 cm above the ground, demonstrating that the detection system works at height as well as at ground level. This type of aerial obstacle would have been completely ignored by a perimeter wire system or even some basic RTK systems.
Behavior with animals and living beings
We conducted several tests with our dog, small garden wildlife (hedgehogs observed during scheduled nighttime sessions), and even with a supervised 6-year-old child. In all cases, the robot maintained a safe distance, generally stopping between 35 and 50 cm from the detected living being. After a 3- to 5-second pause, if the obstacle hadn't moved, the robot made a wide detour, deviating at least 60 cm from its initial path.
We observed that the system seems to identify living beings by their movement, even the slightest (the breathing of a lying animal, the slight swaying of a standing child), then applying a more cautious avoidance protocol than for inanimate objects. This sophistication greatly enhances safety, particularly in homes with young children or free-roaming pets.
One particular case stood out: during an early morning mowing session, a hedgehog curled up in a ball in the middle of the lawn. The robot approached, stopped about 45 cm away, waited 8 seconds, then made a wide detour of about 80 cm before resuming its normal mowing. A few hours later, the hedgehog having left the area, the robot passed through the same zone normally, demonstrating that it does not create a permanent exclusion for a temporarily detected obstacle.
Comparison with RTK systems and cables
Having tested several perimeter wire robots and RTK systems in recent years, we can confidently say that the E18's vision system offers superior flexibility and safety. Perimeter wire robots completely ignore any unmapped obstacles within the defined perimeter, colliding with them until the stop bumper is triggered. RTK systems with ultrasonic obstacle detection offer basic but less precise detection, with frequent false positives (shadows, tall grass mistaken for an obstacle) and a shorter reaction distance.
The E18's 3D vision minimizes false positives thanks to image analysis combined with depth measurement, enabling earlier anticipation and therefore smoother maneuvers. During our testing period, we only recorded two false positives: once when a patch of particularly tall grass (15 cm) was briefly mistaken for a low obstacle, and once when a strong shadow cast by a tree at the end of the day slowed the robot, which ultimately crossed normally after a few seconds of hesitation. False positive rate: less than 1% of detections, excellent performance.
Limitations and borderline cases
Despite its remarkable overall performance, the vision system has some limitations identified during our tests. Very flat obstacles (such as a 2 cm diameter garden hose lying on the ground or an electrical cable) are not always detected, and the robot sometimes passes over them. No damage was observed in our tests, but theoretically, a very flat, fragile object could be struck.
Reflective or transparent surfaces (like glass on the ground or a decorative garden mirror) can confuse vision-based detection. We experimentally placed a large mirror flat on the lawn: the robot approached normally and gently bumped into the frame via the bumper, the reflective surface having confused the system. This is an extremely rare occurrence in normal use, but good to know.
In very low light conditions (late twilight, early dawn), the system loses accuracy. The manufacturer also indicates that the robot automatically returns to its station when the light level becomes insufficient, a useful safety feature. During our tests, this automatic return was triggered approximately 25 minutes after sunset in clear weather, and 40 minutes after in cloudy weather (where residual light is more diffuse).
Ultimately, the E18's real-time obstacle detection system is a major technological achievement, offering a level of safety and flexibility rarely seen in robotic lawnmowers. The ability to navigate safely through cluttered and changing environments, without prior mapping of these temporary obstacles, represents a significant improvement for everyday usability.
Effective rain sensor with adjustable sensitivity ★★★★☆ (17/20)
The rain sensor integrated into the top of the robot provides an important protective function, preventing mowing in wet weather, which can clog the mechanism and degrade cutting quality. We evaluated this system during three rainy periods within our testing phase, with varying intensities and durations.
The sensor is a small, perforated grid located on the top of the robot, allowing water to enter and be detected by an internal capacitive or resistive system. Three sensitivity levels can be set via the app: low, medium, and high.
In high sensitivity mode, the robot detected rain and returned to the charging station after only 8 seconds of exposure during a moderate downpour (estimated at 3 mm/h). This highly responsive setting offers maximum protection but may be overly cautious during light summer drizzle. We observed two unexpected returns during morning fog that would not have warranted stopping the mowing.
At medium sensitivity, detection occurs after approximately 20 to 30 seconds of light to moderate rain. This delay allows the robot to filter out scattered drops or light drizzle without unnecessarily stopping, while also protecting it during heavy downpours. We used this setting for the majority of our tests.
In low sensitivity mode, the robot continued mowing for about 3 minutes in light rain before detecting the moisture and turning back. In heavy rain, the detection occurs more quickly (45 seconds), but this setting seems too permissive because the underside of the robot was already damp and covered in stuck grass when the session ended.
A useful additional feature is the rain pause function. Adjustable from 1 to 4 hours, it defines the waiting time after the rain stops before mowing automatically resumes. On our moderately draining clay-loam soil, we found that a 2-hour delay allowed the lawn to dry sufficiently before resuming mowing without excessive clogging. On very well-draining (sandy) soil, 1 hour might be enough. On heavy, wet soil, 3 to 4 hours would be preferable.
The system generally works well, with 100% detection reliability in our tests (no undetected rain, zero false negatives). The few false positives at high sensitivity are not a major flaw, as they protect the robot. Our only reservation concerns the inability to set finer trigger thresholds than a simple choice between three predefined levels, a software limitation easily corrected by an update.
Questionable value for money compared to the competition ★★★☆☆ (14/20)
At €1499, the Eufy E18 is positioned in the premium segment of autonomous robotic lawnmowers. To assess the relevance of this price, we compared it to two direct competitors representing the two dominant technological approaches: the Worx Landroid Vision WR216E (pure vision technology like the Eufy) and the Husqvarna Automower 430X NERA (EPOS satellite technology).
Versus Worx Landroid Vision WR216E (€1499-€1999 depending on retailer): The Landroid Vision is the most direct technological competitor, sharing the exact same approach: no cables or RTK antennas, based solely on camera vision and artificial intelligence. Offered in a similar or even higher price range (€1499 to €1999 depending on retailer), it boasts a larger capacity of 1600 m² compared to 1200 m² for the E18. The advertised battery life is 70 minutes for the Worx versus 90-110 minutes for the Eufy, reversing the advantage in favor of the tested robot. The Landroid's cutting width reaches 22 cm compared to 20.3 cm for the E18, a minor difference in practice. Both feature ultra-fast, cable-free installation, AI vision obstacle detection, and a comprehensive control app. During our tests, we found that the E18 slightly outperforms the Worx in real-time detection accuracy and responsiveness to moving obstacles, while the Landroid offers a particularly effective Cut-to-Edge function (mowing very close to edges) and self-leveling disc technology to adapt to uneven terrain. The Worx app offers multi-zone management via RFID tags, a feature absent in the Eufy, which requires creating virtual zones on the map. Ultimately, at a similar price (€1499), the E18 wins in battery life and detection, while the Worx excels in maximum capacity and multi-zone flexibility. For a 600-800 m² lawn, the Eufy is a better choice thanks to its longer battery life, reducing the need for recharging. For 1000-1200 m² and larger lawns, the Landroid becomes preferable despite its shorter battery life, as its higher maximum capacity ensures better coverage. Comparative rating: ★★★☆☆
Versus Husqvarna Automower 430X NERA (€3610 and up): The Husqvarna is positioned in a significantly higher price range, more than twice the price of the E18. It uses a radically different technology: the EPOS satellite system with a reference station that creates virtual boundaries using centimeter-level GPS positioning, rather than camera vision. This hybrid approach (satellite + radar for obstacle detection) offers several advantages: a capacity of up to 3200 m² in irregular mode or 4800 m² in parallel lines mode, a battery life of 145 minutes with a 5 Ah battery, slope management up to 50% (compared to 40% for the E18), and radar detection in addition to the positioning system. The Husqvarna X-Line's construction inspires confidence with premium materials, a high-resolution color display (Automower Access), integrated LED headlights, electric cutting height adjustment from 20 to 60 mm, and a 24 cm cutting width. The Automower Connect app offers visual mowing patterns (stripes, checkerboard, triangles), Zone Control for multiple zones with separate settings, Rewild mode (10% of the lawn left unmowed for biodiversity), and home automation integration (Alexa, Google Home, IFTTT). However, the EPOS installation requires a reference station with a clear view of the sky, a requirement not present with the E18. The complete kit (robot + EPOS plug-in + reference station) costs well over €3,600. For a garden under 1000 m² with good Wi-Fi coverage, the E18 at €1499 offers much better value than the Husqvarna, which is only truly worthwhile for gardens over 1500 m² or where there is no stable Wi-Fi connection. For very hilly terrain (40-50% slopes), complex areas with multiple narrow passages and separate zones, or gardens exceeding 2000 m², the Husqvarna justifies its higher price with superior technical capabilities and professional-grade robustness. However, for average-sized home use, the Eufy offers 80% of the features for 40% of the price. Comparative rating: ★★★★☆
Overall, the E18 occupies a coherent middle ground: comparable to the Worx Landroid Vision in terms of price and technology, and significantly more affordable than the high-end Husqvarna NERA. Its main strength lies in the balance between ease of installation (better than the Husqvarna), decent battery life (better than the Worx), and a reasonable price (the same as the Worx, much lower than the Husqvarna). Its main drawback remains its absolute reliance on Wi-Fi and just-sufficient battery life, which is a disadvantage for lawns larger than 700 m². At €1499, the price-performance ratio is justified for gardens of 400 to 700 m² with stable Wi-Fi coverage, but becomes questionable beyond 800 m² where multiple recharges considerably lengthen the mowing time.
Conclusion
Overall rating: 15/20 ★★★☆☆
A true revolution in the world of robotic lawnmowers, the Eufy E18 delivers on the promise of installation as simple as that of a robot vacuum cleaner. In just 10 minutes, without burying any cables or installing a satellite dish, the robot is up and running and maps your garden on its own. This unprecedented accessibility, combined with real-time obstacle detection that is among the most effective on the market and near-professional edging, positions this model as an attractive option for gardens of 400 to 700 m² with stable Wi-Fi coverage. However, the limited battery life of 90-110 minutes is a drawback for areas approaching the advertised 1200 m², requiring multiple recharges that considerably lengthen the overall mowing time. At €1499, compared to the Worx Landroid Vision WR216E (offered in the same price range but with a larger 1600 m² capacity) and especially the Husqvarna Automower 430X NERA (€3610) offering professional capabilities, the E18 occupies a coherent middle ground. It will perfectly suit users who prioritize absolute ease of installation, safety through advanced obstacle detection, and quality construction, while accepting the requirement of essential Wi-Fi coverage and just enough battery life for average-sized lawns. For gardens of 400 to 700 m² with good Wi-Fi coverage, it's an excellent choice combining cutting-edge technology and ease of use. For lawns larger than 800 m², competitors like Worx (for pure Vision) or Husqvarna (for a higher-end option) are worth considering.
Detailed sub-notes:
- Installation and commissioning: ★★★★★ (18/20)
- Automatic mapping: ★★★★☆ (17/20)
- Mobile application: ★★★☆☆ (15/20)
- Mowing quality: ★★★★☆ (16/20)
- Obstacle detection: ★★★★★ (19/20)
- Rain sensor: ★★★★☆ (17/20)
- Electricity consumption: ★★★★☆ (17/20)
- Autonomy: ★★☆☆☆ (12/20)
- Value for money: ★★★☆☆ (14/20)
Eufy E18 Robotic Lawnmower
Also available on Amazon:
Some links on our site are affiliate links. This has no impact on the price you pay, but these commissions allow us to fund our work and continuously improve our tests and comparisons. Thank you for your support!
Our best tests and reviews of robotic lawnmowers
With thirty years of experience in robotic lawnmowers, Husqvarna has built this longevity into the heart of
Robot lawn mowers, Dreame A2 vs. Mammotion Luba Mini Lidar AWD. The robot lawn mower market
Imagine a summer morning. You open your shutters and discover a perfectly mowed garden, without
Today, we are diving into the heart of a real revolution in garden maintenance with Luba