11:30 AM - 11:45 AM
[MTT46-04] Support System for Horticultural Workers Using Augmented Reality
Keywords:Horticulture, Augmented Reality, Information Visualization, Tacit Knowledge, User Interface
In recent years, smart agriculture using ICT has been promoted against the background of decreasing and aging agricultural workers. One of the barriers for new farmers is that many of the cultivation methods used by experienced farmers are based on tacit knowledge acquired through years of experience, and are not clearly explained so that novice farmers can understand. In this study, we focused on greenhouse horticulture, and mainly on eggplant as a field where IT is relatively advanced. The work of trimming unnecessary branches and leaves is done every week, and although it has a great impact on yield and quality, there is a lack of information for new farmers to refer to regarding the parts and the amount to be trimmed. In addition, the long hours of work required to trim are a great burden for farmers in general. It is desirable to be able to intuitively check the information for reference at the work site, because the growth state of each crop varies individually.
In this study, we proposed a method to support trimming branches and leaves using Augmented Reality (AR) technology that superimposes environmental data obtained by sensors on real objects. In this method, the key is how to visualize the tacit knowledge of skilled farmers as explicit knowledge. This objective can be divided into two items. The first is to mathematically convert the tacit knowledge into explicit knowledge. Since there are almost no cultivation manuals at present, we will extract tacit knowledge that can be supported by AR technology. Secondly, we propose a User Interface (UI) for the AR system. The proposed UI is as geometrically simple as possible so that it can be eventually implemented in an AR system that automatically presents information.
This research was conducted with the cooperation of instructors and trainees of an agricultural training facility. Based on the interviews with the farmers, we analyzed the issues and conducted preliminary experiments using a prototype created in Virtual Reality (VR) to further identify the issues. Taking the results into consideration, we proposed the following support methods.
In this study, we assume that the users of the system are farmers in general, new farmers, and short-term employed workers. In the proposed application, the information in the field and the tacit knowledge are superimposed on the actual field through AR glasses. The information includes virtual objects in the form of blocks that indicate plants with large amounts of leaves, objects that indicate leaves with pests and diseases, numerical values of leaf area per ridge (Leaf Area Index), markers that pinpoint areas to be cut, and 3D objects that reproduce the work of a skilled person or display the desired crop shape and messages from the manager. It is a 3D model that serves as a teaching tool.
In order to verify the effectiveness of our method, we first developed a pilot system using AR glasses, and then evaluated the usability of the system by questionnaires to farmers and the consistency of the system with the needs assessment of production sites by semi-structured interviews.
As a result of the verification, two specific tacit knowledge points were identified; they were found to be the range of increase or decrease in leaf volume due to growth and trimming, and the seasonal adjustment of leaf volume. The results of the usability evaluation were outside the acceptable range according to Bangor et al., but relatively good when we focus on UI evaluation.
It is hoped that this method will be used to improve the speed and accuracy of branch and leaf trimming operations, to support better planning of operations, to enrich data-based knowledge, to promote the skill development of new farmers, and to enable the sharing of work goals in numerical form between employees and managers.
In this study, we proposed a method to support trimming branches and leaves using Augmented Reality (AR) technology that superimposes environmental data obtained by sensors on real objects. In this method, the key is how to visualize the tacit knowledge of skilled farmers as explicit knowledge. This objective can be divided into two items. The first is to mathematically convert the tacit knowledge into explicit knowledge. Since there are almost no cultivation manuals at present, we will extract tacit knowledge that can be supported by AR technology. Secondly, we propose a User Interface (UI) for the AR system. The proposed UI is as geometrically simple as possible so that it can be eventually implemented in an AR system that automatically presents information.
This research was conducted with the cooperation of instructors and trainees of an agricultural training facility. Based on the interviews with the farmers, we analyzed the issues and conducted preliminary experiments using a prototype created in Virtual Reality (VR) to further identify the issues. Taking the results into consideration, we proposed the following support methods.
In this study, we assume that the users of the system are farmers in general, new farmers, and short-term employed workers. In the proposed application, the information in the field and the tacit knowledge are superimposed on the actual field through AR glasses. The information includes virtual objects in the form of blocks that indicate plants with large amounts of leaves, objects that indicate leaves with pests and diseases, numerical values of leaf area per ridge (Leaf Area Index), markers that pinpoint areas to be cut, and 3D objects that reproduce the work of a skilled person or display the desired crop shape and messages from the manager. It is a 3D model that serves as a teaching tool.
In order to verify the effectiveness of our method, we first developed a pilot system using AR glasses, and then evaluated the usability of the system by questionnaires to farmers and the consistency of the system with the needs assessment of production sites by semi-structured interviews.
As a result of the verification, two specific tacit knowledge points were identified; they were found to be the range of increase or decrease in leaf volume due to growth and trimming, and the seasonal adjustment of leaf volume. The results of the usability evaluation were outside the acceptable range according to Bangor et al., but relatively good when we focus on UI evaluation.
It is hoped that this method will be used to improve the speed and accuracy of branch and leaf trimming operations, to support better planning of operations, to enrich data-based knowledge, to promote the skill development of new farmers, and to enable the sharing of work goals in numerical form between employees and managers.