The AI Watching Your Back: A New Frontier in Worker Safety

SAN FRANCISCO, CA – August 12, 2026 – In the relentless hum of factories and warehouses, the human body remains the most essential, and vulnerable, piece of equipment. For decades, the effort to protect it from the strain of repetitive motion and heavy lifting has been a slow, manual process. But a new wave of technology is promising to change that, and its latest evolution aims not just to spot danger, but to teach workers how to avoid it in real-time.

Today, San Francisco-based TuMeke launched Coaching Cues, an AI-powered feature that uses computer vision to analyze a worker’s movements and provide instant, actionable feedback. The system, which runs on a simple phone or tablet, can identify a strained reach or a dangerously twisted back from video footage and immediately generate a visual guide—complete with illustrations and video clips of the infraction—to show the worker a safer way. It represents a significant leap in a field rapidly moving from reactive incident reports to predictive, and now prescriptive, safety.

From Clipboards to Computer Vision

For years, ergonomic assessments involved an expert with a clipboard and a stopwatch, manually observing tasks and taking notes. The process was time-consuming, subjective, and the feedback often came days or weeks after the fact. TuMeke, founded in 2019, was part of a new guard of tech firms that sought to automate this process. By using AI to analyze video, their platform creates 3D models of worker movements, generating objective risk scores without the need for wearable sensors or intrusive hardware.

With Coaching Cues, the company is moving beyond mere risk identification. "Identifying risk is important, but safety teams also need to know what to do with that information," said Diwakar Ganesan, Co-founder and CTO at TuMeke, in a statement. "Coaching Cues help connect the analysis directly to action by showing what happened and giving assessors practical guidance they can use with the worker right there in the field." The system can flag common issues like working outside the body’s “power zone,” excessive back bending, or one-armed lifts, turning a risk score into a teachable moment.

This immediate feedback loop complements the company’s broader “AI Recommendations,” which suggest larger-scale fixes like changing workstation height or process design. "Ergonomic risk is rarely solved by one intervention," Ganesan added. "Sometimes the immediate opportunity is helping someone adjust how they perform a movement. Other times the job itself needs to change. Our goal is to give teams the insight and tools to address both."

The Economic Imperative

The push for such technology is fueled by a staggering economic reality: musculoskeletal disorders (MSDs) cost U.S. employers an estimated $54 billion annually in workers' compensation, lost productivity, and medical expenses. For companies like Siemens, New Balance, and Cargill—all TuMeke clients—reducing these injuries is not just an ethical obligation, but a powerful lever for financial performance.

Early adopters of the technology report significant benefits. One safety manager at a large manufacturing firm noted that the platform allows them to "easily identify and prioritize the highest risk tasks faster," turning weeks of analysis into a matter of hours. Another praised the shift from subjective debate to data-driven action, stating the system provides "objective, comparable insights" that allow experts to focus on solutions rather than manual data gathering.

This value proposition has attracted serious investment, including a $10 million Series A funding round led by Intel Capital in late 2023, earmarked for product development. It has also opened up novel business models. In a strategic partnership, TuMeke’s platform is offered at no cost to certain business insurance customers, turning the technology into a risk-mitigation tool for the insurers themselves and embedding it directly into the financial ecosystem of workplace safety.

An Unblinking Eye: The Human Questions of AI Monitoring

While the promise of a safer workplace is compelling, the deployment of AI-powered computer vision raises profound questions about privacy and surveillance. As cameras and algorithms become permanent fixtures on the factory floor, where is the line between a helpful coach and an unblinking digital supervisor? The technology’s ability to constantly monitor, analyze, and score human movement introduces a new dynamic into the employer-employee relationship.

Legislators are grappling with this new reality. In the United States, a patchwork of state laws governs employee monitoring, with regulations like the California Privacy Rights Act (CPRA) granting employees new rights over the data collected about them. Internationally, the framework is stricter. The European Union’s AI Act, which began taking effect in 2025, classifies AI systems used for worker monitoring as “high-risk,” imposing stringent transparency and compliance obligations on employers.

TuMeke states its platform features a "privacy-first design that anonymizes visual data" and uses "human-in-the-loop AI technology." The goal, they argue, is to analyze the task, not the person. Yet the potential for this data to be used for performance management, or for algorithmic biases to creep into assessments, remains a critical concern for labor advocates. The challenge lies in ensuring these powerful tools are used for empowerment and safety, not for punitive surveillance.

A Crowded Field

TuMeke is not alone in this pursuit. The AI-powered ergonomics market is projected to swell to over $8.5 billion by 2033, and a growing number of competitors are vying for a piece of it. Companies like SoterAI and Visionify Inc. offer similar computer vision solutions, while large Environmental, Health, and Safety (EHS) software providers like VelocityEHS and Cority are integrating AI into their broader platforms. The race is on to provide the most accurate, scalable, and actionable insights.

In this crowded landscape, TuMeke is betting that its focus on a wearable-free, 3D assessment platform, now enhanced with an integrated coaching component, will be a key differentiator. The introduction of Coaching Cues signals a clear strategy: to own the entire workflow from risk detection to behavioral correction, making ergonomics faster, more accessible, and more effective. The ultimate success of this technology, however, will depend not only on the sophistication of its algorithms, but on the trust and transparency with which it is deployed among the very workers it is designed to protect.