Ramping challenges in a critical manufacturing segment

The power of modern robotics is in the ability to handle the unpredictable using hallmark human strengths – intelligence and real-time physical feedback. While the manufacturing space has benefited for decades from traditional, rigid automation, processes demanding enhanced dexterity or adaptability to variable conditions continue to require human workers.

Surface finishing tasks such as sanding, grinding, polishing and painting fall into this category. Demand for such specialized labor is high – and in markets like the United States, a manufacturing resurgence and re-shoring trends are creating tailwinds. In the US, per conservative estimates, there are several hundred thousand workers employed in finishing tasks alone.

However, there is a big problem. These roles are physically demanding and require training and skill. They are performed in dust-filled and toxic environments that threaten health and safety. As a result of high churn, manufacturers are perpetually short on qualified staff and cannot meet demand. And at the end of it all, production results are far from ideal: humans are prone to inconsistent output quality. Given these myriad challenges, the space is crying out for adaptable automation paradigms.

Enter GrayMatter Robotics – robust, AI-driven flexible automation

We at NGP Capital are delighted therefore to announce our Series B investment today in GrayMatter Robotics. The Los Angeles-based startup has developed a robotics platform for high-mix manufacturing, which is characterized by 1) a large variety of parts (usually in low volumes), and 2) a high degree of variability within each SKU, meaning even the same nominal parts have individual differences due to imprecise production. Because in high-mix environments the concept of a manufacturing “line” is informal, there is even variability in how parts are moved around, fixtured, and oriented.

GrayMatter tackles all this uncertainty using standardized, off-the-shelf industrial robotics hardware paired with a curated sensor suite and powers it with a proprietary computing engine. The system autonomously assesses the size, orientation and geometric complexities of the target part – no CAD model required – via a 3D imaging scan in under 60 seconds. Factoring in material properties, it then determines the optimal trajectory of the robot arm and end effector settings (e.g. translational speed, RPM, applied force) – all without any programming.

As the robot executes its work plan, it relies on vision and force sensors for real-time feedback. This gives it the ability to dynamically course-correct with new information. The end result is a machine that can think (and learn) on the fly, taking on more than 90% of the difficult and dangerous work that humans typically perform.

As its brain-inspired name suggests, most of GrayMatter’s secret sauce lies in software intelligence, represented by a fusion of models serving what can best be described as “adaptable AI”. The team concluded early that a combination of reinforcement learning with real-world heuristics and physics-based inputs – versus end-to-end AI – was the best approach. Theirs is a perfect example of resisting the urge to force-fit a technology to a problem, but rather assembling the optimal set of tools without bias to achieve the best practical outcome.

Customers realize immediate benefits from deployment of GrayMatter systems. The robot clocks at 2-4X human speed, boosting production output and revenue. Finish quality is substantially higher and more consistent, and consumables usage drops by up to 50%. The UI/UX is beautifully simple and designed for operators without prior robotics expertise.

Critically, affordability is not an issue. The Robotics as a Service (RaaS) business model, one that achieved scale in the hands of Rick Faulk and his team at my past investment Locus Robotics, has been instrumental in democratizing access. The recurring payment approach meets customers where they are.

The value proposition has resonated. In the past year, GrayMatter executed an exciting commercialization growth spurt across a national base. Customers have signed up quickly across a broad spectrum of industries and product types – from aerospace to maritime to sports & recreation.

Backing the team and tech – personal connections and thesis synergies

The success is not a surprise. When I first met Ariyan, the co-founder and CEO, more than two years ago, the product was still in development. However, I was drawn to its technical sophistication – a direct benefit of the founders’ academic backgrounds. More importantly, what stood out was how deeply the team was cycling in the field with customers to immerse themselves in everyday operational problems and iterate on product improvements.

That approach was a direct hit on our robotics thesis at NGP, which fits into our broader firm focus on digital-physical convergence. It was also the foundation that led to subsequent check-off of other key points for us, namely: (i) creating a data moat backed by proprietary domain knowledge (ii) achieving 90%+ uptime with minimal human intervention (iii) delivering immediate and compelling customer ROI and (iv) productizing go-to-market and deployment steps, aiming to drop-in and steer clear of bespoke implementations.

Additionally, we are excited by the power of platforms versus standalone products. Within an initial robotics use case, once a base AI “brain” is established, hardware integration dialed in, and threshold performance demonstrated – itself a challenging milestone – extension to first adjacent and later more distant applications becomes possible. GrayMatter is executing on that playbook. It has developed a series of dexterous, contact-based solutions and is launching a new product set that automates painting and coating.

While the team is technically a knockout, this breadth of vision marks them as truly special. They have an authentic passion to transform an industry and improve the lives of everyday working families. With that motivation and deeply differentiated technology in hand, we believe they are positioned to revolutionize many key manufacturing segments.

As participants in GrayMatter’s Series B financing, we at NGP are thrilled to partner up with Ariyan, his co-founders Brual and SK and the rest of the team on the ambitious journey to bring flexible robotics into manufacturing and set the industry on a new course.

______

We continue to be excited by the vast opportunities at the convergence of manufacturing, AI and automation. If you are building a company in this space, I would love to hear from you. And if you are interested in joining the fast-growing team at GrayMatter, please contact them here.

‍