CASE STUDY — SAFETY
Improving safety in mining
A human centred approach
*More details of the project can be discussed in person because of client confidentiality.
Project Overview
Mining operations present some of the most complex safety challenges in industry. Despite rigorous protocols, unsafe behaviours and conditions persist — driven by habit, time pressure, and systemic barriers to reporting. This project applied a human-centred design approach to understand the root causes of unsafe acts and co-create solutions with the frontline workers themselves.
Unsafe Conditions
Physical hazards in the environment — inadequate lighting, unstable structures, faulty equipment — that workers encounter daily and often normalise over time.
Unsafe Behaviors
Actions or inactions by workers that deviate from established safety procedures — skipping checks, bypassing equipment, and under-reporting near-miss incidents.
Challenges
- 01
Workers skip safety checks under time pressure, creating invisible compliance gaps.
- 02
Near-miss incidents are systematically under-reported due to fear of blame and a cumbersome reporting process.
- 03
Existing safety tools were designed by compliance teams without input from frontline workers, resulting in poor adoption.
- 04
Supervisors lacked real-time visibility into hazard patterns across shifts and locations.
Research Process
01
Ethnographic Research
02
Quantitative Assessment
03
Focus Group Discussions
04
Ideations & Workshops
Ethnographic Research
1. In-depth Interviews
We conducted 24 interviews with frontline workers, supervisors, and safety officers across three sites. Sessions were held on-site to build trust and reveal the unspoken norms that shape daily behaviour.
Interview documentation
2. Ride Alongs
Researchers accompanied workers through full shifts — from pre-task briefings to post-shift handovers. This surfaced the real decision points where safety shortcuts occur and why.
Ride-along documentation
Quantitative Assessment
1. Quantitative Research
A structured survey of 320 workers across all sites measured safety attitudes, tool adoption rates, and incident frequency perception. Data was segmented by role, experience, and shift pattern.
2. Technical Assessment
Audited the existing safety reporting system — logging flows, completion rates, error patterns, and time-to-submit. Revealed an average of 14 steps to file a near-miss report.
Focus Group Discussions
Five focus groups — each with 8–10 workers of mixed seniority — were facilitated to validate findings from the ethnographic phase and co-develop problem framings.
Focus group session photo
Research Outputs
Identified 12 core behavioural triggers that lead to unsafe acts, mapped to specific task contexts and environmental conditions.
Created an end-to-end safety incident journey map — from hazard recognition through reporting to resolution — revealing 6 critical failure points.
Defined 5 design principles for safety-critical interfaces, grounded in the cognitive load constraints of the operational environment.
Workshops and Ideations
Three co-design workshops brought together workers, supervisors, safety officers, and digital product teams. Participants sketched, prototyped, and stress-tested concepts.
Workshop session 1
Workshop session 2
Prototype testing
Ideation Outputs
Redesigned safety reporting workflow — reduced from 14 steps to 3 with photo capture and voice-to-text support for low-literacy contexts.
Visual risk dashboard concept enabling supervisors to monitor hazard hotspots and near-miss trends across shifts in real time.
Outcome and Impact
47% reduction in unreported near-miss incidents within 6 months of launching the new reporting tool.
30% → 78% increase in daily safety check completion rates across all three sites.
200+ workers trained on the new digital safety tools with an average onboarding time of under 20 minutes.
Top-3 hazard categories identified and addressed through targeted physical interventions co-designed with site engineers.
Supervisor visibility transformed from weekly paper reports to a live dashboard updated every 15 minutes.
Zero lost-time injuries in the 90 days following full rollout — the longest injury-free period in the site's recorded history.
Recognised by the client as a template for rollout across their 8 other operational sites globally.