Recognition of lawn information for mowing robots

In this paper, the method for recognition of lawn information in mowing robots (Robomower) is proposed. The mowing robots will be expected to work in open areas such as parks. To work in an open area, the robots must keep the safety of the public. It is necessary to collect surrounding information adequately. The Robomower is a small and low-power robot, which uses GPS information to get coordinates of its current position. It has many sensors: GPS, terrestrial magnetism direction sensor, ultrasonic wave sensors, tactile bumper switches etc. However, the robot did not have the means to measure the height of the grass. Reflection type sensors and image analysis system are not effective for this task. We need to recognize the lawn to complete mowing. We succeeded in measuring grass height with a simple, cheap sensor. The method is to estimate the height of the turf by using a transmission type Photointerrupter. The effectiveness of the sensor was confirmed by experimentation. It is necessary to collect information in a real environment to decide the action of the robot. The purpose of past research o was an efficiency gain of the entire work, and mowing in a real environment was not a purpose. The entire efficiency is high' is more useful for the cooperation working of robots than 'actual mowing work'. The acquisition means of appropriate environmental information is necessary for 'actual mowing work'. However, a turf sensor was developed to acquire necessary information on the lawn. The information on the lawn necessary for the mowing work is the places of height turf areas. To acquire the lawn information, transmission type's sensor is developed for recognizing the lawn environments. In this method, a simple, cheap sensor made the situation acquisition possible. The output result is shown below.