From Reactive to Proactive: How Computer Vision AI is Transforming Workplace Safety

The traditional approach to workplace safety has always been reactive... waiting for incidents to occur before taking action. But what if we could predict and prevent workplace injuries before they happen? That's exactly what R/VISION, in partnership with ACC, is achieving through groundbreaking computer vision AI technology that's transforming how New Zealand businesses approach workplace safety.

The Scale of the Challenge

The statistics paint a sobering picture of workplace safety in New Zealand's manufacturing sector. Manufacturing accounts for 19% of total work-related ACC claims, with 240,400 work days lost to injuries annually in the sector alone. The economic burden is staggering - $1.23 billion annually in manufacturing injuries across the New Zealand economy.

As one safety professional put it: "A safe workplace is integral to a healthy, engaged workforce and a thriving economy. But there are thousands of things to focus on in any worksite environment. Even the most vigilant person can't see, record, react or analyse everything that's happening around them."

This fundamental limitation of human observation has driven the need for innovative solutions that can monitor, detect, and predict safety risks in real-time.

Enter R/VISION: The AI Safety Revolution

R/VISION represents a paradigm shift from reactive incident management to proactive risk prevention. Using advanced computer vision AI models, the system continuously monitors workplace environments through existing CCTV infrastructure, identifying potential safety risks before they become incidents.

The technology works by applying sophisticated AI models that track what's happening in chosen environments, providing near real-time event detection that empowers a continuous improvement cycle based on the Plan-Do-Check-Act (PDCA) methodology:

Plan: Identify your risk profile and safety monitoring requirements

‍Do: Let R/VISION gather data and analyse trends automatically

‍Check: Review near-real-time alerts that prompt immediate team action

‍Act: Use feedback loops to continuously improve the data and responses

Breakthrough AI Models in Action

The partnership with ACC has enabled the development of cutting-edge AI models specifically designed for New Zealand workplace conditions:

Ergonomic Risk Detection

The system can identify three critical types of ergonomic risks that commonly lead to workplace injuries:

1. Torso Rotation: Detecting when workers twist their bodies in ways that could cause strain
2. Bending: Identifying potentially harmful bending postures
3. Reaching: Monitoring for overextension that could lead to injury

These AI models consistently achieve 95% detection accuracy across all monitored safety risks during real-world trials, representing a fundamental shift in how workplace safety is managed.

Fall and Slip/Trip Detection

Beyond ergonomic monitoring, the system also detects potential fall hazards and slip/trip risks, providing comprehensive coverage of the most common workplace injury scenarios.

Exclusion Zone Monitoring

The AI can identify when workers enter restricted or dangerous areas, providing immediate alerts to prevent potential incidents.

Building Trust Through Shared Governance

One of the most significant achievements of the R/VISION implementation has been worker acceptance and cultural integration. Rather than being viewed as surveillance, frontline teams have embraced R/VISION as a "safety guardian."

This success is built on a foundation of shared governance with key industry partners including:

• Downer: Leading infrastructure and transport company
• EMA: Employers' and Manufacturers' Association
• +IMPAC: Workplace safety specialists
• Te Rūnanga O Toa Rangatira: Providing crucial Māori perspective and cultural guidance
• First Steps Health, Safety, Environment: Bringing additional expertise to implementation

The system demonstrates respect for Māori data sovereignty principles and worker privacy through features like automatic face blurring, fostering a proactive safety culture where workers are incentivised to maintain compliance standards rather than feeling monitored.

The Path Forward: Scaling Innovation

The success has created a blueprint for nationwide deployment across New Zealand. R/VISION's modular architecture enables rapid scaling across operations, with each additional site benefiting from AI models already trained on New Zealand-specific conditions.

The system's integration capabilities mean organisations can leverage existing CCTV infrastructure, maximising return on previous technology investments while gaining cutting-edge safety capabilities.

ACC's continued involvement through the $935,000 Workplace Injury Prevention Grant has validated the solution's cross-industry applicability, with expansion into manufacturing demonstrating the technology's versatility beyond its original infrastructure applications.

Looking Ahead: The Future of Proactive Safety

This partnership has catalysed sector-wide adoption of AI safety technologies, establishing R/VISION as the go-to platform for AI-powered workplace safety. The ongoing development of new AI models for enhanced PPE monitoring and advanced risk detection ensures customers remain at the forefront of safety technology.

The collaboration continues to drive platform enhancement, with continuous development ensuring sustained competitive advantage and operational excellence for organisations committed to protecting their most valuable asset—their people.

As we move from reactive to proactive safety management, the R/VISION and ACC partnership demonstrates how technology, when implemented thoughtfully and with genuine respect for worker concerns and cultural values, can create safer workplaces while building stronger, more engaged teams.

The future of workplace safety isn't about replacing human judgment—it's about augmenting human capability with AI that never stops watching, learning, and protecting the people who matter most.

‍