<img src="https://ml.globenewswire.com/Resource/Download/81cc8890-badf-45a7-8552-574790347b28?size=2" nborder="0" align="left" hspace="5" vspace="5" title="GrayMatter Robotics logo.png" alt="GrayMatter Robotics"/><ul><li>Manufacturers across aerospace, defense, specialty vehicles, and shipbuilding are converging on the same constraint: surface finishing work varies too much, part to part, for fixed-path automation to handle</li><li>GrayMatter Robotics’ proprietary data regime, ATLAS, comprises real-world surface finishing data accumulated across 30 million square feet of surface area, including multiple materials, industries, environments, and synchronized sensing modalities </li><li>Physical AI-based autonomous finishing systems deliver up to 12 times the throughput of skilled manual labor, according to GrayMatter Robotics’ deployment data</li><li>Part programming drops from weeks to under five minutes with no manual reprogramming required via GrayMatter Robotics’ Factory SuperIntelligence (FSI)</li></ul> Carson, CA, Aug. 06, 2026 (GLOBE NEWSWIRE) -- <a href="https://www.globenewswire.com/Tracker?data=-tnFy7BnKmGjZ9UXBStaK0DhXXGHwz5KIYOeX_hIo1MRnQ9yn3CI-LKq7g_Cgkcx5Sj9hIwq7EO7AE_jgryTGXeqvnYyXObVjLma72D5u-xC0dQFk29l1pLyqmlurCkyK6-pmaOwHiA-fAoh4t9C3uIUfT0-PxbAySoswvXUsdzuLsyFAQ9hr3rR0qJAhUEvKBCZ6ZnnYgfhkYblnvXI0OsOGBHpMREvJMc-r6Ki6To=" rel="nofollow" target="_blank" title="Physical AI">Physical AI</a> is the manufacturing industry's answer to a wall that high-mix production lines keep hitting with surface finishing: fixed-path robots work reliably when every part is identical and break down the moment surface condition, geometry, material, or tool condition varies from one unit to the next. That constraint has kept sanding, grinding, blasting, and polishing largely manual across industries where part variation is the norm, from aerospace components to shipbuilding to specialty vehicles. GrayMatter Robotics, the Physical AI company building <a href="https://www.globenewswire.com/Tracker?data=ckveOs7KU-1GTnzMWBh_gTo8zjEShQIxOIHbd2I9ou9yhIWJSzZqVbnALkQoPy93ujGLSkYRHTrtJ2pNY6tyCLMlX-pkp7U0xmS7FrfG8JMIMEflMMuRPvgfbpxw6ANxXLqLx9mbKBj4myxBfoRI_w==" rel="nofollow" target="_blank" title="Factory SuperIntelligence (FSI)">Factory SuperIntelligence (FSI)</a> for industrial automation, has deployed that architecture across 20+ industries including aerospace, defense, specialty vehicles, and shipbuilding, processing over 30 million square feet of surface area.  "What surprises people is how much a finishing cell in aerospace benefits from what we learned in shipbuilding or specialty vehicles from what we learned in aerospace. Physical AI carries across industries even when the parts look nothing alike, and that's the leverage ATLAS gives us. A new deployment starts with the accumulated experience of each one that came before it," said Ariyan Kabir, Co-Founder &amp; CEO, GrayMatter Robotics, who holds a PhD in robotics and artificial intelligence.  Key Facts: <ul><li>GrayMatter Robotics has raised $70.4 million in total funding and operates a 100,000-square-foot AI Robotics Innovation Center </li><li>In defense MRO applications, GrayMatter Robotics' Physical AI finishing systems deliver 10 times the productivity of skilled manual labor</li><li>On average, a 90% reduction in ergonomically challenging manufacturing processes across GrayMatter Robotics deployments</li><li>Consumable waste is reduced by 30 to 50% through consistent force and pressure control across Physical AI finishing deployments</li></ul> Question: What Is Physical AI and How Does It Differ from New-age AI Software Systems? Answer: Physical AI refers to AI systems that operate in and learn from the physical world. New-age AI software systems power large language models and digital assistants. They are trained on internet-scale text and image data, reasoning about language and patterns. Physical AI is trained on physical interaction data: force, pressure, torque, contact geometry, surface topology, and material response under real operating conditions.  The practical consequence of this distinction is significant for manufacturing because new-age AI software systems cannot tell a robotic system how much pressure to apply to a titanium aerospace component or how to adjust its tool path when a composite skin deviates from its nominal dimensions. Physical AI systems develop this capability through accumulated contact experience, learning from millions of finishing operations across parts and conditions. Question: How Do Autonomous Robots Sense, Reason, and Act in Factory Environments? Answer: Autonomous robots powered by Physical AI operate through a continuous real-time loop that mirrors how an experienced human operator applies judgment to a finishing task. <table align="left" style="border-collapse: collapse; width:100%; border-collapse:collapse ;"><tr><td style="text-align: center ; vertical-align: middle; vertical-align: top ; ">Phase</td><td style="text-align: center ; vertical-align: middle; vertical-align: top ; ">What Happens</td><td style="text-align: center ; vertical-align: middle; vertical-align: top ; ">How Physical AI Executes It</td></tr><tr><td style="vertical-align: top ; ">Sense</td><td style="vertical-align: top ; ">Multi-modal sensor arrays scan the actual part before and during operation, building a real-time picture of geometry, surface condition, and material properties from physical contact</td><td style="vertical-align: top ; ">Vision, force torque, acoustic, and surface sensing systems work simultaneously, reading each part exactly as it arrives, variation included</td></tr><tr><td style="vertical-align: top ; ">Reason</td><td style="vertical-align: top ; "><a href="https://www.globenewswire.com/Tracker?data=EoglAwF5A-Fr9krLBtIGPKSmcLbCtUuT860Ua76bdCVcwkqMg-BRYU8GypuvfyL3vnOUaJoe81uOZFCnKrAIErvIpsNRFjRwOyvZeJ96ZaCdrCZNviRKDbpRG2nmlHDl" rel="nofollow" target="_blank" title="Process Intelligence">Process Intelligence</a>, GrayMatter Robotics' term for the learned understanding of how tools, media, and workpiece materials co-evolve during process execution, where complex contact produces controlled material change (removal, deposition, deformation, surface modification) rather than serving as a positioning constraint. Developed through ATLAS rather than pre-programmed physics models</td><td style="vertical-align: top ; ">The system determines the correct force, pressure, speed, angle, and tool path for the specific part in front of it, drawing on accumulated experience across similar materials and geometries</td></tr><tr><td style="vertical-align: top ; ">Act</td><td style="vertical-align: top ; ">The robotic system executes the finishing strategy in real time, adjusting thousands of times per second as conditions change</td><td style="vertical-align: top ; ">When a surface irregularity is detected, force control compensates immediately. When tool wear shifts contact pressure, the system recalibrates. Output holds consistent across part number 1 and part number 500, regardless of operator shift or workforce experience level</td></tr></table> What Is the Difference Between Physical AI and Traditional Industrial Robotics? Traditional industrial robots follow pre-defined paths optimized for high-volume, low-variation production. That approach works well for standardized parts but requires per-part programming that can take weeks per new geometry and cannot adapt to real surface conditions or material variation during operation. Across defense, aerospace, and specialty vehicle production in particular, where part-to-part variation is the rule rather than the exception, that limitation has kept fixed-path autonomy out of surface finishing entirely. Physical AI systems can address this by learning from accumulated contact data across real manufacturing environments, generating finishing strategies from physical interaction rather than pre-programmed rules, and adapting in real time to the actual part in front of them. This process determines whether autonomy is viable for high-mix manufacturing or limited to high-volume production lines. Question: What Is Factory SuperIntelligence? Answer: Factory SuperIntelligence (FSI) is GrayMatter Robotics' intelligence platform for industrial automation, purpose-built for physical manufacturing environments. FSI combines proprietary AI models, domain agents, and process optimization across factory operations, delivering rapid deployment and reconfiguration capabilities. It brings Physical AI and new-age AI software systems together for manufacturing across industries, environments, materials, geometries, and applications, functioning as the AI layer that runs the factory floor rather than automating individual manufacturing processes in isolation. About GrayMatter Robotics Headquartered in Carson, California, GrayMatter Robotics is building Factory SuperIntelligence (FSI) that powers the autonomous factories of the future. Founded in 2020, the company develops Physical AI technologies and deploys autonomous factories that handle complex, high-mix tool-manipulation applications such as surface preparation, coating, and inspection processes across some of the most demanding production environments in the world, delivering up to 12x the throughput of skilled manual labor and up to a 95% reduction in rework. Its air-gapped, edge-deployed architecture ensures full data sovereignty for defense and enterprise-critical operations. To date, GrayMatter Robotics has processed over 30 million square feet of surface area across 20+ industries, serving customers in aerospace, defense, shipbuilding, specialty vehicles, and consumer products. The company is on a mission to reindustrialize American manufacturing and bolster our National Security, bridge the gap between demand and capacity of our industrial base, and ensure the industrial resilience the nation depends on. For more information, visit <a href="https://www.globenewswire.com/Tracker?data=85FcmKV3Bcvix0KCfXu2vQxo3GToRo7koCkQpTc7vQNs6Rq-UfScHGk0IKeHgwyIcotR37ibXSfllwPVvPyoKIi3bItpSzdiQhC2IDjcMjT8KftxNOwZSXCI-LbFchBb" rel="nofollow" target="_blank" title="graymatter-robotics.com">graymatter-robotics.com</a>. <pre>CONTACT: Sarah Evans
Head of PR, Zen Media
sarah@zenmedia.com</pre> <img alt="" src='https://ml.globenewswire.com/media/ZmM3YmQwOGMtMTc4ZC00MjJiLWEzZWUtOTU3Yjc5NjIyZGVhLTUwMDE2MDUwMS0yMDI2LTA4LTA2LWVu/tiny/GrayMatter-Robotics.png' referrerpolicy='no-referrer-when-downgrade' />

GlobeNewswire Inc.

What Is Physical AI, and How Do Autonomous Robots Work in Factories?

Latest news from the world of business and finance – both local and global

MONEY FM 89.3 is Singapore's first and only Business and Personal Finance station, featuring local and international business and financial news.

MONEY FM 89.3

Discover newsletters from our DJs on LinkedIn