Adaptive multimodal safety systems with incident localization and visualization

A non-transitory computer-readable storage medium including instructions that, when executed by one or more processors, cause the one or more processors to perform operations comprising:

• (a) detecting an event based on time-series data collected from a wearable device, wherein the time-series data comprises both motion data and physiological data about a subject;
• (b) obtaining location data and time data about the event, wherein the location data is provided by a real-time locating component;
• (c) identifying and selecting one or more imaging devices from a plurality of imaging devices based at least in part on the location data and a field of view of individual ones of the plurality of imaging devices;
• (d) determining, by a first machine learning model that has been trained to detect event types of events, a type of the event;
• (e) extracting, based at least in part on the time data about the event and the type of the event, a visual data segment from visual data captured by the one or more imaging devices identified and selected in (c);
• and
• (f) processing the visual data segment by a second machine learning model that has been trained to identify causes of events, to identify a cause of the event.