2021 Fiscal Year Research-status Report
Development of Smart Viticulture System Using Autonomous Robots
| Project/Area Number |
21K14117
|
|---|
| Research Institution | Kitami Institute of Technology |
|---|
Principal Investigator |
ラワンカル アビジート 北見工業大学, 工学部, 准教授 (70802594)
|
|---|
| Project Period (FY) |
2021-04-01 – 2024-03-31
|
| Keywords | Robotics / Autonomous Robots / Smart Agriculture / Artificial Intelligence / Computer Vision |
|---|
| Outline of Annual Research Achievements |
This research aims to realize smart viticulture using autonomous robots and artificial intelligence. There are several objectives of the project which are expected to be achieved over three years. In the first year (FY2021), we developed the Simultaneous Localization and Mapping (SLAM) module for vineyards using Lidar and camera sensors. A simulator for vineyards was developed and the developed SLAM algorithm was tested. The algorithm was further tested in actual vineyard with real robots. Moreover, a navigation module was developed and tested in actual vineyard. The navigation was successfully achieved without using GPS sensor, using only Lidar and camera sensors. The navigation module could autonomously navigate the robot in the lanes of the vineyard while smoothly avoiding the static and dynamic obstacles. In addition, several datasets of grapes and weeds in dynamically changing scenarios were collected using multiple sensors. The recognition module of grapes is currently being developed. The results of the SLAM module and the navigation module for vineyards have been presented in international journal, international conferences, and domestic conferences. Valuable feedback was achieved from these conferences and reviewer comments. The SLAM and navigational module will serve as the basis for the next targets.
|
| Current Status of Research Progress |
Current Status of Research Progress
2: Research has progressed on the whole more than it was originally planned.
Reason
This research aims to realize smart viticulture using autonomous robots and artificial intelligence. The research progressed as planned initially. Although COVID pandemic impacted the delivery of robot, the research was progressed first with simulation environment and later in actual vineyard. The Simultaneous Localization and Mapping module and the navigation module forms the core of the proposed research. Vineyard poses several challenges due to uneven terrain and lack of GPS data. In this regard, the SLAM and navigation module were successfully completed as planned. We conducted an evaluated study to determine appropriate sensors and actuators to be used for the actual robot and based on this results the tests on real robots will be determined for the next fiscal year.
|
| Strategy for Future Research Activity |
For the FY2022, the plan is to develop the grape and weed recognition module with highly accurate localization. The robot mechanism for grape harvesting and weed removal will also be tested. First, the simulation model will be developed. Later, real-hardware will be developed. Actual tests in field are also planned. Additional sensors that were not available in the previous FY are planned to be procured and tested for data acquisition. For realizing the long-term grape recognition, monitoring, navigation for mobile robots, tests in several dynamic scenarios including low-light conditions (several days to weeks) are planned with continuous data acquisition. At the later stage of this FY, the goal is to train a deep learning framework that is robust to severe changes in the environment. At different stages of the research, the results will be published in conference and technical journals.
|
| Causes of Carryover |
Due to the COVID impact several robot components necessary for the execution of the research could not be obtained. In the next fiscal year, we would like to proceed with the purchasing of the necessary items.
|
Research Products
(6 results)
