Agentic AI, robotics, and automation are revolutionizing the manufacturing industry as the industry faces a shortage of skilled workers.
Over half a million manufacturing jobs remain unfilled. This gap is expected to surge to 2.1 million jobs by 2030, according to a report by Deloitte and TheManufacturing Institute. Businesses are already experiencing difficulties. Taiwan Semiconductor Manufacturing Company (TSMC) cited a shortage of skilled workers as the primary reason for the delay of its $65 billion plant in Arizona. Those delays could impact American tech firms, which may become more apparent next year.
DeepHow’s Smart Know-How Companion empowers workers to ramp up faster, perform smarter, and stay safer. Inspired by NVIDIA’s visual language models, DeepHow expanded its AI capabilities to provide task guidance, verification and compliance for frontline workers.
These innovations aren't about replacing people, they're about enabling them.
Key elements of this expanded functionality include:
- Visual Verification: Use AI to compare a worker's recorded task execution against a gold-standard reference video or procedure.
- Pinpoint Differences: Highlights step-by-step differences or deviations, showing exactly where something was done differently or incorrectly.
- Task Accuracy & Safety: By visually flagging errors or skipped steps, correct mistakes before they become costly or dangerous.
- Continuous Improvement: Supervisors can use it to provide targeted coaching and feedback that accelerates skill development and reduces rework.
- Audit and Compliance: Create a visual record of completed tasks to support quality audits, process compliance, and training documentation.
DeepHow's100+ enterprise customers create content 90% faster, improve knowledge by 50% and onboard 77% faster with their AI-driven, Smart Know-How Companion.
Anheuser-Busch InBev uses DeepHow across its breweries worldwide, including sites in US, China, India, Vietnam, and SouthKorea. Faced with urgent upskilling needs and a retiring workforce, they replaced outdated, paper-based training with DeepHow's AI-powered video training. As a result, they reduced onboarding time by 80%, improved training consistency, and preserved critical knowledge across their frontline teams.
ArcelorMittal turned to DeepHow and uncovered inconsistencies between written procedures and actual workflows. Packaging these changes with other operational improvements into DeepHow’s AI Videos and Skills Management, resulted in a 49% reduction in delayed incidents and a 78% reduction in safety incidents.
Stanley Black & Decker captured over 30 years of expertise to transfer critical knowledge. They useDeepHow to engage Millennial and Gen Z workers, which allows them to simplify job requirements and promote career development.
DeepHow’s platform uses NVIDIA’s GPUto accelerate AI processing, enabling video analysis, transcription, and smart segmentation. DeepHow also leverages the NVIDIA Blueprint for video search and summarization from Metropolis and is excited to explore NVIDIA’s new reasoning vision language model — Cosmos Reason. DeepHow plans to leverage Cosmos Reason’s ability to understand physical environments, including factories, production lines, field service installations, and repairs.
Today's workforce, particularly Millennials and Generation Z, expects on-demand learning with visual guidance and real-time feedback to build confidence on the job. When manufacturers do this, they will see jobs completed faster, safer, and correctly, thereby increasing the productivity and proficiency of their labor force.
As NVIDIA redefines what's possible with Visual AI, DeepHow is focused on putting those capabilities in the hands of the people who build, assemble, repair, and maintain the world around us.
