整合AIOT防護系統設計應用於機車騎乘

根據公路總局公開資料機車遭竊案件居高不下，且機車騎士行駛於道路狹窄及車輛動線規劃不完善之路面，在沒有機車車體保護情況下，身處易遭碰撞受重傷的風險中。為解決上述問題，本研究結合物聯網（Internet of Things，IoT）、4G連網，以及影像辨識技術，從而建置AIoT智慧物聯網系統。本系統利用微型單板電腦來進行實作，將AIoT智慧物聯網分為三個子系統：1.透過影像辨識進行車輛識別進行預警，達到車輛逼近安全距離時，警示使用者或低於安全距離，即將產生碰撞即時通知緊急聯絡人；2.車輛若遭竊使用者可以可遠端遙控模擬車輛故障狀態，致使宵小無法繼續進行騎乘行為；3.若使用者行駛之車輛遭碰撞，將透過SMS（Short Message Service）傳送至使用者設定之緊急聯絡人GPS（Global Positioning System）定位資訊，抑或者車輛遭竊也將透過SMS通知使用者，使用者可以進行即時定位，獲取宵小目前所在位置通知使用者。本研究期望建立更完善的人身安全保護與機車財產保護措施，來降低駕駛意外與延遲送醫的機率及保護使用者車輛。

關鍵字

AIoT ；物聯網；影像辨識

並列摘要

The motorcycles theft would result in the weak anti-theft device of motorcycles and the fact that stolen motorcycles are not easy to be seized. According to the statistics of the number of motorcycles registrations published by the General Administration of Highways, the cases of motorcycles theft remained still high. Moreover, when many motorcycles were driving on roads with narrow roads and imperfect motorcycles movement planning, the riders of motorcycles would be at risk of being seriously injured in collisions because they could be not protected by the motorcycles' covers. In order to solve the above problems, this research combined the Internet of Things (IoT) and artificial intelligence (Artificial Intelligence, AI) technologies to build a smart AIoT system. The AIoT system was implemented using a micro single-board computer and divided into three subsystems. This AIoT system was applied the image recognition technology for vehicle identification for early warning, used to obtain the longitude and latitude positioning data of GPS, and been remote shutdown. This study expects to establish better personal safety protection and motorcycles anti-theft measures to reduce the probability of driving accidents and delayed hospitalization and protect the users' motorcycles.