Connect the NWT Device
Use the Android tablet and OTG to complete serial detection, permission prompts and device handshake so the on-site hardware chain is verified before measurement.
Workflow Narrative
Break the measurement chain into four interpretable stages, from device connection to record storage.
The public workflow page explains only the technical path that can be shared openly: device connection, normalization, live sweeping, force conversion and data sync. It helps clinicians and partners understand the product without exposing real clinical data.
Measured Flow
Realtime Protocol
步骤
4
把准备、采集、识别与留存整理成更易理解的流程分镜。
提示
3
把异常提示、重测规则与下载说明放到显性位置，减少理解成本。
Workflow Steps
Every public step corresponds to an actual measurement action inside the clinical system.
围绕临床现场的使用节奏组织流程卡片，让准备、采集、识别与留存形成一条清晰路径。
Connect
Ready
先确认设备连接与工作状态，避免流程一开始就中断。
Baseline
Stable
把基线采集单独立起来，形成可信的数据起点。
Realtime
Loop
扫频、换算与同步形成连续闭环，而不是孤立步骤。
Run Normalization
Capture the empty-load Normalize baseline first, then compare and correct later live sweeps against that baseline to stabilize the measurement origin.
Sweep Live and Locate the Valley
Collect spectrum curves continuously at a fixed step size, display the S-parameter trace in real time and automatically identify the resonance valley.
Convert Force and Sync Records
Convert resonance frequency into orthodontic force with the six-parameter sensor model, save locally first and then synchronize to Django on demand.
Protocol Mapping
The workflow page should explain the underlying protocol and algorithms, not only the visual flow.
For a public site, the highest-value content is not a full screenshot of the internal UI, but a structured explanation of handshake packets, sweep data, Normalize correction and force modeling.
协议层不再只是说明文字，而是流程、波形和规则的可视化组合。
后续可直接替换为扫频曲线、谷值定位或设备通讯示意素材。
Protocol to Interface
把准备、识别与结果呈现组织成同一块视觉容器，让流程理解更完整。
Serial Handshake
Serial Handshake
The same handshake header and command padding rules from the desktop NWT protocol are preserved for migration consistency.
Sweep Data
Sweep Data
Each point is read as 4 bytes, with the first 2 bytes in big-endian order used for amplitude conversion.
Spectrum Correction
Spectrum Correction
The empty-load baseline is subtracted from live data to build the corrected spectrum used for valley detection.
Force Model
Force Model
A two-term exponential model with a, b, c, d, mean and std converts resonance frequency into force.
Public Boundaries
The workflow can be public, but privacy and clinical risk boundaries must remain explicit.
公开流程页不仅说明步骤，也要明确何时需要重新确认连接、重做基线与获取最新资料。
Explain methods, not private cases
The public workflow page presents abstract stages such as connection, normalization, sweep, conversion and sync, without showing any real patient screens.
Warnings must be visible
Serial disconnections, validation failures, abnormal fluctuation and baseline expiry are all foregrounded inside the business system, and the public narrative should explain these risks clearly too.
Separate the public site from the business system
The website explains capabilities, workflows and downloads, while real treatment records, patient information and force data remain inside authorized terminals.