AI/machine learning

Vision Analytics Can Decrease HSE Risk and Improve Worksite Efficiency

Vision analytics is being used to extract insight information from video, with data inferred from existing cameras used to create a monitoring dashboard where supervisors can receive alerts at the worksite level or drill down to specific events.

208773_Fig.jpg
Vision analytics can be used to detect and track personnel activity.
Source: Paper SPE 208773

Computer vision is no longer an emerging technology. It is being actively used for accident prevention as well as operation optimization by many companies. For instance, in the aviation industry, airport gate security cameras are being repurposed to allow continuous process improvement with computer vision. Self-driving cars are another major benefactor of the advances in this technology.

The aim of the work described in this paper was to

  • Evaluate several machine-learning technologies to identify the best fit use cases
  • Execute a proof-of-concept project to gauge feasibility of the chosen methods
  • Create dashboards
  • Build work flows that make the data consumable with existing and emerging technologies
  • Test the ability to track key performance indicators (KPIs)
  • Access the value for potential benefits

This paper describes the practical aspects vision analytics, part of a larger effort to create a smart factory for maintenance shops. The technology could be use for wellsites as well. Data work flows are designed to transition to systems that will automatically use this information in the future. The combination of video feed data and data from business systems results in process tracking, automated safety compliance detection, and continuous improvement of operational KPIs.
The goal of using these computer vision tools is to reduce operational risks by increasing compliance with the use of personal protective equipment (PPE) and enforcing high-risk safety zones. Creating a system for tracking people and equipment together with object detection and tracking using tags will create automated time tracking for each maintenance operation and thereby automating operational KPI reporting.

Automated KPIs will give management continuous metrics for process improvement. Creating a movement and activity heatmap will allow for ooptimization of workshop layout as well as process optimization to increase overall efficiency. This project was focused on several areas of interest as an initial proof of concept.

  • PPE compliance
  • Object detection and monitoring
  • Risk zone monitoring
  • Spaghetti movement diagram
  • Asset utilization
  • People position and activity
  • Bay utilization
  • Tripping alerts
  • Lifting compliance

These fall into two basic categories: safety and efficiency. There is also a larger list of use cases in which several technology areas could be utilized, such as data analytics from sensors, automated tracking with global positioning systems, and time-series prognostics.
Download the complete paper from SPE’s Health, Safety, Environment, and Sustainability Technical Discipline page for free until 8 February.

Find paper SPE 208773 on OnePetro here.