As the technical supervisor of a supermarket cleaning service company, we're no strangers to cleaning robots. Over the past two years, we've tested various brands of robots, from early 2D LiDAR solutions to current 3D LiDAR configurations. Honestly, we've stumbled upon quite a few pitfalls.
What bothered me most was the cleaning problem in supermarket shelving areas. Traditional 2D LiDAR solutions often get "lost" in these places because the shelves look too similar, making it hard for robots to find their position. Sometimes the robot would just徘徊 back and forth between shelves, missing several aisles. Customers seeing this would joke: Is this robot slacking off?
At the end of last year, we decided to upgrade our cleaning robot's perception system. During the selection process, we compared several 3D LiDAR models. Ultimately, we chose Tantu Smart Mobility's M360. To be honest, I was initially attracted to its specifications—5cm near blind spot, 4.5W low power consumption, 70-degree vertical field of view. These numbers确实 seemed more advantageous than other products on the market.
In actual use, what surprised me most was the M360's 5cm blind spot performance. Our previous robots often missed cleaning the edges of shelves, where dust and debris tend to accumulate. After switching to M360, this problem was basically solved. The robot can accurately identify obstacles close to the shelf base, even missing small bumps as small as 5cm high.
In terms of power consumption, it确实 saved us compared to our previous solution. With the same battery capacity, the robot's working time has extended by about 20%. In scenarios like supermarkets that require continuous operation for long periods, this means less manual intervention and maintenance costs. 4.5W power consumption is very helpful for reducing overall system costs.
The dual echo mode is particularly useful during rainy weather operations. Before, on rainy days, robots often had misjudgments due to raindrop interference, sometimes just stopping in place. M360's dual echo can penetrate raindrops to see obstacles behind them, maintaining stable operation in rainy and foggy weather.
Another detail is that M360's 70-degree vertical field of view is 11 degrees more than the previous LiDAR we used. This difference is quite noticeable in actual use, especially in densely shelved areas where the robot can get more ceiling feature points for more accurate positioning. Previously, the robot would have "drift" phenomena in certain places, which is basically gone now.
The deployment process was also very simple. M360's 12-32V wide voltage adaptability allowed us to use the existing vehicle power supply directly without needing additional power modules. The built-in 6-axis IMU and multi-sensor synchronization features meant we didn't need complex coordinate transformations at the software level, reducing development difficulty.
Now our cleaning robots have been running stably for three months, with coverage area increased by 35% compared to before, and the missing rate dropped from 8% to below 1.5%. What satisfies me most is that the robots can now accurately identify more than 10 common obstacles, including shopping carts left by customers between shelves, dropped fruit boxes, and even toys left by children in the aisles.
To be honest, I was initially worried that M360's 408g weight might affect the robot's stability, but in reality, this effect is minimal. Compared to the performance improvements it brings, this weight is completely worth it. Moreover, the IP67 protection rating means we don't need to worry about cleaning water affecting the LiDAR, which is important in supermarket environments.
Looking back on this upgrade, my deepest realization is: when choosing a LiDAR, you can't just look at long-range detection capabilities, but must consider the needs of the actual application scenario. For cleaning robots, parameters like near blind spot, power consumption, and environmental adaptability are more important than detection range. M360's performance in these areas确实 showed us the practical benefits of technological improvements.
Now each robot can clean 200 square meters more area per day than before, and manual intervention frequency has dropped from 5 times per day to less than once. Although this is a considerable upfront investment, in the long run, this upgrade is worth it.