WiFi 鏡頭看見人和探頭，WiFi 探頭回報是否真的接觸；你網路裡的邊緣節點把肌群輪廓疊到畫面上。MyoNav APP 帶你掃描，AI 助理把口語描述對應到身體部位。 A WiFi camera sees the person and the probe, a WiFi probe reports real contact, and an edge node on your own network overlays muscle outlines on the view. The MyoNav app guides the sweep; the AI assistant maps plain-language descriptions to body regions.
實線在區網內；虛線才會連到雲端，而且不帶影像與人體骨架點。Solid lines stay on your network; only the dashed lines reach the cloud — without video or body landmarks.
每一層沒連上時，上一層照樣運作：沒有鏡頭就用 APP 單機導航，沒有雲端就在區網內完成疊圖。If a layer is missing, the one above keeps working: no camera, use the app standalone; no cloud, the overlay still runs on your network.
固定在腳架上，拍下整個人與探頭；只負責串流，不在鏡頭端做任何判斷。On a tripod, it films the whole person and the probe. It only streams — no decisions are made on the camera.
三個壓力感測器判斷是否真的貼在皮膚上，連同電量與訊號強度一起無線回報。Three pressure sensors tell whether it is really on the skin, reported wirelessly with battery and signal strength.
放在同一個網路裡的電腦：辨識人體姿勢與探頭位置，把肌群輪廓對到畫面上，再判斷這一刻能不能計入覆蓋。A computer on the same network finds body pose and probe position, aligns muscle outlines to the view, and decides whether this moment can count toward coverage.
MyoNav APP 顯示 3D 肌群、掃描路線與疊圖；登入後同步紀錄，並可使用 AI 助理。The MyoNav app shows 3D muscles, sweep paths and the overlay; signed-in users sync records and can use the AI assistant.
只有在鏡頭看得清楚、而且探頭確實接觸時，覆蓋才會往前走。Coverage only advances when the camera sees clearly and the probe is really in contact.
鏡頭把畫面、探頭把壓力與電量，經 WiFi 送到邊緣節點。The camera sends video and the probe sends pressure and battery over WiFi to the edge node.
區網LOCAL辨識人體姿勢與探頭上的標籤，算出探頭在身上的位置。Body pose and the probe's tag are detected to place the probe on the body.
區網LOCAL把選定肌群的輪廓對到畫面上，並給一個信心燈號。The chosen muscle outline is aligned to the view with a confidence badge.
區網LOCAL壓力感測確認貼合；按太大力或感測飽和時，改標為「未知」。Pressure sensors confirm contact; pressing too hard or saturating marks it as “unknown”.
區網LOCALAPP 即時畫出掃過的表面；登入時，只有摘要會同步到雲端。The app paints covered surface live; when signed in, only a summary syncs to the cloud.
APP → 雲端APP → CLOUD看得清楚、位置可信：顯示肌群輪廓，探頭接觸時計入覆蓋。Clear view, trusted position: outline shown, coverage counts while the probe is in contact.
光線不足、姿勢晃動或轉身：只給提示，不計入覆蓋，並告訴你原因。Low light, movement or turned away: hints only, no coverage, with the reason shown.
鏡頭斷線、人或探頭離開畫面：超過 1 秒沒有新畫面就停止疊圖，不沿用舊結果。Camera dropped, person or probe out of view: after one second without a new frame the overlay stops — stale results are never reused.
用真實的邊緣節點程式，搭配模擬的鏡頭畫面與探頭訊號，在手機與平板模擬器上實際操作 APP。The real edge-node software, fed with simulated camera frames and probe signals, driven through the app on phone and tablet emulators.
以上為 2026 年 9 月的軟體模擬測試結果，使用合成畫面與模擬訊號，不是臨床量測，也不代表實機表現；實機驗證完成後會更新。 Software simulation results from September 2026 using synthetic frames and simulated signals. Not clinical measurements and not a statement of hardware performance; will be updated after hardware validation.
邊緣節點在你的網路裡，官網看不到它；配對後的連線狀態請在 APP 內查看。The edge node lives on your network, so this site can't see it — check its connection inside the app after pairing.
目前採邀請制封閉測試。留下信箱，開放名額時會寄試用連結與安裝說明給你。We're in an invitation-only beta. Leave your email and we'll send a link and install guide when a spot opens.