SoterCoach Starter: Reports – Soter Help Center

To help you better understand and manage movement risks, SoterAI provides two downloadable reports: the Hazards Hour-by-Hour Breakdown, which summarizes hazardous activity by time of day for easier trend analysis, and the Raw Events Data Report, which details every individual movement event captured.

📊 Hazards Hour-by-Hour Breakdown (CSV)

The Hazards Hour-by-Hour Breakdown report provides a high-level summary of hazardous movement activity, grouped by one-hour intervals. This report allows you to quickly identify time periods with increased ergonomic risk and understand how different hazard types are distributed throughout the workday.

🕐 What Does This Report Show?

Each row in this CSV file represents a one-hour time window on a given date. The report aggregates both the total number of movement events and a breakdown of hazardous events by type.

Use this report to:

Pinpoint high-risk hours across a shift
Compare hazard trends over time
Monitor the impact of interventions or coaching

⚠️ Understanding Hazard Counts

Each hazard category shows how many events triggered that specific classification during the hour. A single event can be counted in one hazard type only, following the hierarchy used by the device.

The field "Hazardous Events" reflects the total number of unique events during the hour that were classified as hazardous (i.e., where at least one of the five hazard flags was 1).

📁 Field Descriptions

Field	Description
`date`	Calendar date of the recorded events (YYYY/MM/DD).
`hour_range`	One-hour time window in which events were grouped (in 24-hour format).
`total_events`	Total number of movement events recorded during the hour.
`hazardous_events`	Count of events during the hour where at least one hazard flag was `1`.
`awkward_static_posture`	Number of events classified as hazardous due to awkward static posture.
`repetition`	Number of events classified as hazardous due to repetition.
`poor_bending`	Number of events classified as hazardous due to poor bending.
`back_twisting`	Number of events classified as hazardous due to back twisting.
`intense_bending`	Number of events classified as hazardous due to intense bending.

ℹ️ Each hazard category is mutually exclusive — an event is only counted under one type, based on system classification logic.

🧾 Sample Template Preview

Below is a sample layout of the report format. You can use this to familiarize yourself with how the data is displayed

Hazards Hour-by-Hour Breakdown Template

📊 Raw Events Data Report (CSV)

The Raw Events Data Report provides a row-by-row view of every movement event recorded by a SoterCoach device. This data can be used to audit movement quality, identify risk patterns, and conduct deeper analysis of workplace ergonomics.

🔍 What Is an Event?

Each row in the report corresponds to a single movement event captured by the sensor. Events may include bending, twisting, or holding a static posture. Every movement is time-stamped and enriched with sensor data and hazard classifications.

⚠️ What Counts as a Hazardous Event?

Not every event is hazardous.

An event is marked hazardous if any of the following five hazard flags are set to 1:

Awkward Static Posture
Repetition
Poor Bending
Back Twisting
Intense Bending

If all five hazard fields are 0, the event is non-hazardous and represents routine, low-risk motion.

📁 File Format and Field Descriptions

The file is in .csv format and can be opened in Excel, Google Sheets, or any spreadsheet tool. Here's a breakdown of the core columns included:

✅ Hazard Flags

Field	Description
`awkward_static`	Binary flag for hazardous static posture. Set to 1 if `average_bend_angle` ≥ 60 and `time_static` ≥ 20 seconds. If set to 1, all other hazards are 0.
`repetition`	Binary flag for hazardous repetition. Set to 1 if `risk_enabled` is true for current and any event in the past 60 seconds.
`poor_bending`	Binary flag for poor bending. Set to 1 if `bend_angle` ≥ 85°.
`back_twisting`	Binary flag for back twisting. Set to 1 if `bend_angle` ≥ 50 and `lower_back_twist_angle` ≥ 70, or if `bend_angle` ≥ 75 and `twist_angle` ≥ 35.
`intense_bending`	Binary flag for intense bending. Set to 1 if `bend_angle` ≥ 85 and machine learning model identifies intensity.

✅ Angle Measurements

Field	Description
`bend_angle`	Peak forward bending angle (in degrees) during the event.
`twist_angle`	Peak twisting angle (in degrees) of the torso during the event.
`average_bend_angle`	Mean bending angle sustained throughout the event (in degrees).
`average_twist_angle`	Mean twisting angle sustained throughout the event (in degrees).

✅ Time and Meta

Field	Description
`timestamp`	Unix timestamp marking the start of the event.
`time_start`	Human-readable format of the event start time.

✅ Other Fields

All additional fields not listed above are labeled as:

Technical field used for backend processing.

These fields are used internally to support hazard detection and analytics.

🧾 Template Preview