Smart Plant Watering System: A Mobile Robotic Solution for Precise Plant Hydration in Indoor Environments
Abstract
Serenity, responsibility, and mindfulness are but a few of the prominent positive states of mind that get activated when we are taking care of our beloved houseplants. Going away for a long business trip or vacation, occasional lapses of memory due to old age, and physical limitations of people are some of the primary reasons why many indoor plants wither down. Currently, there are a few solutions that can detect dehydration levels and water plants accordingly. Such solutions mostly utilize IoT sensors and Arduino/Raspberry Pi-controlled watering devices. Although these solutions are effective, they require a lot of bulky setups that are not easy to install, take up considerable space, and are difficult to transport if we decide to change the location of the plant. House plants ideally can be allocated around the house: on shelves, corners of the room, in different rooms, and even near windows. For pinpointing such dynamic locations, using a small-scale positioning system that utilizes UWB (Ultra-Wide Band) which has the best wall-penetrating signal strength and can give us location precision of around 10 cm is proposed. The use of a robot-arm with computer vision capabilities will allow for detecting, reaching, and watering the plants with great accuracy. The indoor arrangement of each household is unique and to tackle vehicle navigation along with obstacle avoidance SLAM (Simultaneous Localization and Mapping) along with path planning will be implemented. This allows the robot-vehicle to map the dynamic environment and traverse to the destination of dehydrated plants. To identify the dehydrated plants, thresholds of soil moisture based on the type of plant are set up, and the real-time moisture will be recorded via soil moisture sensors and obtained through their API. To demonstrate the effectiveness of the solution, several tests in indoor scenarios were conducted and how successful the robot performed in each setting was noted down.