Shell invests in safety with Azure, AI, and machine vision

Summary

Shell is leveraging computer vision and machine learning to enhance safety at its retail sites, which serve 30 million customers globally. By analyzing CCTV footage in near real-time, they can identify safety risks like vehicle incidents and theft, allowing for timely interventions. The integration of Azure cloud services with advanced technologies like TensorFlow enables effective processing and continuous improvement of safety measures. This innovative approach not only aims to enhance operational efficiency and reduce costs but also provides staff with vital insights, ultimately transforming how Shell conducts its business.

Takeaways

• 😀 Retail is a crucial component of Shell, serving 30 million customers daily across over 70 markets.
• 🔒 Shell's safety goal is tied to Goal Zero, ensuring safe visits for both customers and staff at all retail sites.
• 🚧 Retail locations face various safety risks, including vehicle incidents, theft, smoking, and improper refueling.
• 👁️ Enhanced video analysis through computer vision can help detect and intervene in potential safety risks in near real-time.
• 💻 Traditional analysis of CCTV footage is labor-intensive; computer vision streamlines this process significantly.
• ☁️ Azure cloud services facilitate the integration of machine learning models for real-time data processing and alerts.
• 📊 Deep learning technologies help filter and identify important events in footage for further review and algorithm training.
• 🕒 Continuous, real-time data feeds will enhance operational safety compared to traditional post-event analysis.
• 👀 The insights gained provide additional oversight for service staff, improving safety measures.
• 🚀 Early pilot stages of the technology show potential for significant improvements in effectiveness, cost reduction, and safety enhancements.

Q & A

What is Shell's primary focus in retail?

Shell's primary focus in retail is to ensure safety at their sites, aligning with their Goal Zero initiative to provide a safe environment for customers and staff.

What safety risks are associated with Shell's retail sites?

Safety risks at Shell's retail sites include vehicle collisions, theft, smoking, and incorrect refueling behaviors.

Why is video footage analysis important for Shell's retail operations?

Video footage analysis is important because it allows Shell to monitor safety incidents in real-time, facilitating timely interventions and enhancing overall safety.

What challenges did Shell face with manual CCTV footage analysis?

Shell faced challenges due to the labor-intensive process of manually analyzing extensive CCTV footage captured daily at their retail sites.

How is computer vision being utilized in Shell's operations?

Computer vision is used to analyze video footage and identify events of interest in near real-time, which helps improve safety and operational efficiency.

What technologies does Shell leverage for video analysis?

Shell leverages Azure cloud services, including IoT Hub and Azure Databricks, along with open-source technologies like TensorFlow and Kafka.

What is the role of deep learning in Shell's safety strategy?

Deep learning helps filter video footage to identify relevant safety incidents and uses labeled footage to continuously train and improve algorithms.

What benefits does near real-time data provide to Shell's service champions?

Near real-time data provides service champions with enhanced situational awareness, acting as a 'third pair of eyes' to monitor safety effectively.

What stage is Shell currently in regarding the implementation of this technology?

Shell is in the early stages of a pilot project to leverage this technology at scale and explore its potential to transform business operations.

What opportunities does Shell see in implementing this technology?

Shell sees opportunities to improve effectiveness, reduce costs, and enhance safety through the broader implementation of this technology.