$200K, 4 Universities, 71 Workers: What the NSC's Latest MSD Research Actually Found

Musculoskeletal disorders have been the largest category of workplace injury in the U.S. for 25 consecutive years, costing employers billions annually in workers' compensation, lost productivity, and turnover. The National Safety Council's MSD Solutions Lab just published findings from its 2024–2025 Research to Solutions grant program — nearly $200,000 across four university teams — and the results are a must-read for any EHS leader building an evidence-based ergonomics program. Risk & Insurance reported the full findings on July 13.

Four Studies, Four Technology Approaches

Each grant targeted a different technology-driven approach to reducing MSD risk in real work environments — not labs. Here's what the data showed:

• NC State (AR): $49,999 — Meta Quest 3 headset overlaying color-coded 3D reach zones. 20 workers across 3 facilities (agricultural machinery, pharma). Task completion: 9–41 seconds. Workers rated it intuitive; best for onboarding and workstation reassessment.
• Wichita State (Exoskeletons): $48,467 — Hilti EXO-S passive arm-support exoskeleton tested across 4 construction trades (drywall, plumbing, sheet metal, electrical). 8 workers, up to 3 weeks. Helpful for overhead tasks; comfort and ease-of-use rated high. But: no statistically significant change in discomfort, strength, or joint mobility.
• Oregon State (Smartphone AI): $49,999 — OpenCap open-source motion capture via smartphones to estimate lumbar spine compression during lifting. 21 participants, 919 lifting sequences. ML model achieved 9% NRMSE vs. 12% for conventional models. Two phones worked as well as three.
• Virginia Tech (AI Bias): $50,000 — Examined sex-based bias in wearable sensor ergonomic assessments. 22 participants. Conventional ML models showed clear sex-based accuracy disparities. A debiasing autoencoder cut mean absolute error to 3.42 kg vs. 4.89 kg for Random Forest.

The Exoskeleton Finding That Should Give You Pause

The Wichita State result is the most important takeaway for anyone running an exoskeleton pilot: workers liked the device, but objective measurements showed no significant physical improvement. Comfort ratings and intention-to-use stayed high, yet shoulder strength, joint mobility, and body-part discomfort didn't change. This is a critical reminder that worker perception ≠ clinical outcome — and that pilot programs need both subjective and objective measurement protocols.

The findings from the 2024-2025 grant recipients demonstrate that emerging technologies can help organizations better understand workplace risks, engage workers and implement targeted solutions that improve safety outcomes.

What to Do With This

For EHS and ergonomics teams: The smartphone-based spinal load assessment from Oregon State is the most immediately actionable finding. If you're still relying on expensive lab equipment for ergonomic assessments, a two-phone OpenCap setup at 9% accuracy is a game-changer for field assessments. The AI bias finding from Virginia Tech is a warning: if you're deploying wearable sensor assessments, audit your models for sex-based accuracy gaps before scaling.

Sources: Risk & Insurance — Research Finds AR Tools, Exoskeletons and AI Systems Show Promise (Jul 13, 2026); National Safety Council MSD Solutions Lab