Incident Analytics

Context and Prerequisites for Automation

Incident management is a basic need for any developing HSE system. However, the transition from paper logs to digital analytics is often accompanied by staff resistance, data distortion, and information loss during the transfer stage. The presentation details the practical experience of building a transparent system for collecting incident feedback, implemented by a team led by Vera Konshina at logistics facilities.

Defining Metrics and Event Classification

For correct data analysis, it is necessary to clearly define exactly which events are subject to accounting. The speaker examines an approach where incidents are divided into four key categories: accidents, micro-injuries, dangerous situations (previously classified as near misses), and non-occupational injuries or illnesses in the workplace.

Special attention is paid to recording non-occupational injuries and deterioration of employee well-being. Although these events are not directly related to production factors, recording them helps protect the employer from hidden occupational injuries and more accurately assess the overall condition of the staff. Each incident category has its own set of metrics: from basic (date, time, facility) to specific (type of unsafe behavior, first aid kit usage).

Data Collection Tools: Accessibility and Transparency

The key problem in recording micro-injuries and dangerous situations is the reluctance of employees to report them. To overcome this barrier, a maximally simplified way of registering events is proposed. QR codes leading to a short electronic form are placed in all operational areas.

• Anonymity and terminology adaptation: The form allows data to be submitted without authorization. Complex English terms have been replaced with clear phrasing, and most fields are implemented as drop-down lists to minimize manual text entry.
• Integration with related processes: Data collection is not limited to employee initiative alone. The system automatically generates requests when scanning a QR code on a first aid kit (a signal of a micro-injury and the need for restocking) or when dangerous maneuvers are detected by built-in systems of material handling equipment.

Task Routing and Execution Control

The speaker demonstrates by example how a completed form is transformed into a task (ticket) in the internal tracking system. The processing is automated: the system independently assigns a responsible HSE specialist depending on the location and involves related departments (fire safety, facility management).

Integration with messengers is used for rapid response. Notifications about critical incidents are instantly sent to the duty group, allowing investigations to begin without delay. At the same time, the request initiator (if the form is not filled out anonymously) sees all stages of processing their appeal, which builds trust in the system and stimulates further staff engagement.

Data Analytics and Non-Obvious Insights

The accumulated data array makes it possible to identify hidden patterns. For example, an analysis of incident times showed that the peak of incident recording occurs during daytime hours, rather than night shifts. This is due to the fact that more walk-throughs are conducted during the day and more hazardous conditions are identified, whereas at night, control weakens. This insight led to a revision of the site audit schedule.

The system also allows tracking the efficiency of the HSE department itself (ticket closure times) and identifying systemic problems, such as the prolonged lack of resolutions for violations by contractors. In the future, the accumulated statistics will be directly integrated into the occupational risk assessment process.

What you will learn from this webinar:

• How to set up incident classification so that data is suitable for automated analytics?
• What tools help overcome staff resistance and motivate employees to report micro-injuries without financial incentives?
• How to link the use of workplace first aid kits with the automatic registration of micro-injuries?
• How to build a transparent algorithm for working with tickets from the moment a request is submitted to the implementation of corrective actions?
• How to use incident statistics to identify blind spots in the production control system?