農家や小規模な自治体が導入可能でかつ害獣による農作物被害対策に有用な「IoT技術を用いた鳥獣捕獲罠用通知装置」を設計、検証を行った。1年目の成果であるWebブラウザーベースのクラウドシステムを用いたメールやSNS（LINE）によって罠の状況を通知する罠監視システムに加えて、罠の振動検知に利用するsens`itデバイスの振動検知の仕組みを調査した。sens`itは防水性を持ちつつオンライン上で5段階の感度の調節を行えるLPWA(Low Power Wide Area)通信モジュールである。sens`itに加速度測定器を装着し、各感度において10回程度振動を与え、各振動における振動検知の状況と加速度の大きさのデータを取得してイノシシが捕獲された際の振動の大きさと比較して最適な設定値を見出した。さらに、Edge式AIカメラモジュールを用いた害獣判別・通知システムも設計、試作した。学習データを作成する際にアノテーションという処理を行うが、アノテーションの方法が異なるモデルを2つ用意し、50枚のイノシシの写真に対する各モデルの認識率を測定した。その上で、LPWA通信とWebサービス(IFTTT)で、振動検知と画像認識の通知がGoogleスプレッドシートを介してslackに送信されるよう設計した。このシステムは、低コストLPWA通信とオープンクラウドなどの利用により端末費用や通信費用が安価で、くくり罠や箱罠への設置台数の飛躍的増大が可能になり野生鳥獣数の生息域拡大を減少に寄与する可能性を有している。今後、この捕獲通知システムを搭載したくくり罠の数を増やしデータの蓄積と通知パターンの分析を行い、捕獲通知の確実性を向上させていく。今年度も、コロナの影響のため会合や罠設置に自粛と制限がかかったが、将来的には行政・県機関との連携を再開して、野生鳥獣による離農や耕作放棄の問題に向き合って行きたい。今年度は食料被害の対策に効果的なイノシシ対策から始めたが、鹿など他の鳥獣へ拡大していく事で「IoT技術を用いた鳥獣捕獲罠用通知装置と農業・地域連携の推進」による持続可能な社会実現を目指していきたい。
A "notification device for bird and beast traps using IoT technology," has been designed and verified in this research. This system can be introduced by farmers and small municipalities and is useful for countermeasures against damage to crops caused by vermin. In addition to the trap monitoring system that notifies trap status by e-mail and SNS (LINE) using a web browser-based cloud system, which was the result of the first year, I investigated the vibration detection mechanism of a sens`it device used for trap vibration detection. Sens`it is waterproof and can be used online. I attached an acceleration measuring device to sensor`it, and gave it about 10 vibrations at each sensitivity, and obtained data on the status of vibration detection and the magnitude of acceleration at each vibration, which was used for the detection of wild boars when they were captured. The optimal setting values were found by comparing the magnitude of vibration. In addition, a vermin identification and notification system using an Edge-type AI camera module was also designed and prototyped. Two models with different annotation methods were prepared, and the recognition rate of each model was measured for 50 photos of wild boars. The system was then designed to use LPWA communication and a web service (IFTTT) to send notifications of vibration detection and image recognition to slack via a Google spreadsheet. This system has the potential to contribute to the reduction of wild animals by dramatically increasing the number of traps and box traps installed, thereby reducing their habitat expansion. In the future, I plan to increase the number of trap sets equipped with this capture notification system, accumulate data, analyze notification patterns, and improve the reliability of capture notifications. This fiscal year, too, I had to refrain from holding meetings and setting traps due to the impact of Corona. In the future, I would like to resume cooperation with administrative and prefectural agencies to address the problem of farmers leaving their farms and abandoning cultivation due to wild birds and animals. This year, I started with wild boars, which are effective in preventing food damage, but I would like to expand the scope to include deer and other birds and beasts in order to realize a sustainable society through "promotion of IoT technology-based notification devices for bird and beast traps and collaboration between agriculture and local communities.
|
