A very important manifestation of the so-called technological progress liberating human productivity is that some tedious and meaningless labor can be done by machines, such as sweeping the floor. Many people now have robot cleaners in their homes. They greatly reduce the time people spend doing housework. However, not all sweeping robots can do well. For example, avoiding obstacles is a good indicator for reviewing the sweeping robot. If it is not done well, the cleaning efficiency and effect will be greatly reduced.
Also Read: Xiaomi Stone Sweeping Robot Won European Hardware Award 2020
Avoiding obstacles approaches of a sweeping robot
The sweeping robot products currently on the market mainly have the following solutions in avoiding obstacles: Lidar obstacle avoidance, 3D structured light obstacle avoidance, 3D ToF obstacle avoidance, monocular vision obstacle avoidance, and binocular vision avoidance.
Let’s briefly analyze. In fact, radar obstacle avoidance, 3D structured light obstacle avoidance, and 3D ToF obstacle avoidance can be collected into one category. They all perceive environmental characteristics by measuring the surrounding depth information. Then, they achieve the effect of avoiding obstacles.
With the exception of radar, the cost and development difficulty of these several technical solutions are relatively low. But they all have one common shortcoming, that is, the perception accuracy is low, it is difficult to identify the characteristics of objects, and it is impossible to actively think about avoidance strategies.
Monocular and binocular obstacle avoidance are methods that use AI image recognition. Among them, monocular recognition is relatively weak in development difficulty and recognition effect, while binocular recognition, the skill to measure depth information, can identify object features. So it is basically the best and most practical one of the current obstacle avoidance technologies of sweeping robots. But it has a strong development threshold.
How does Roborock T7 Pro perform?
Because of this, there are relatively few robot sweeping products on the market that use this technology. The Roborock T7 Pro is one of them.
The Roborock T7 Pro’s AI binocular object recognition + laser planning algorithm combination of technical solutions can achieve unattended fully automated cleaning.
The AI binocular vision module of the Roborock T7 Pro is composed of two 5MP 120° wide-angle cameras, infrared fill light, rechargeable sensor and other components. It is also equipped with a Qualcomm 8-core processor. Through binocular stereo vision and AI objects recognition technology, it realizes intelligent obstacle avoidance function.
At the same time, it is also equipped with a radar laser module and supports LDS laser scanning technology. It can scan 360 degrees and obtain environmental information in real time. Thereby, it can quickly, accurately and comprehensively scann the home terrain.
With the blessing of these two technologies, the Roborock T7 Pro can almost recognize obstacles like ordinary humans sweeping the floor.
Actual experience of this sweeping robot
For example, in our test, we put a shoe on the cleaning route of the T7 Pro. After the sweeping robot detects the obstacle, it decelerates, and then the camera is turned on. After the obstacle is recognized as a shoe, the T7 Pro It will be bypassed directly. At the same time, on the map on the mobile app side, you can see that T7 Pro has synchronized the information of identifying a shoe.
When the T7 Pro encounters a patch panel, it will recognize it in time and do similar obstacle avoidance processing.
The Stone Roborock T7 Pro has realized the recognition of 7 types of common obstacles, including shoes, bases, weight scales, patch panels, thread balls, excrement, socks, dustpans, U-shaped chairs and other types will be added later.
If there is a special object on the ground that the T7 Pro cannot currently recognize. Say, it can’t recognize a photographic light stand, it will not rush forward, but slow down. Thus, it will try to scan forward and feel the shape of the obstacle.
In addition, the Roborock T7 Pro has also introduced an LED infrared fill light in the recognition module. So it can safely avoid obstacles in a dark environment.
In general, the Roborock T7 Pro is very suitable for families with more complex ground environments and geek users who are interested in fresh technology. If you meet these two conditions, then T7 Pro will obviously be an ideal purchase choice.