Our Investment in UnitX

By Varun Jain and Jon Hennessy

The Economic Cost of Quality

In November of 2020, General Motors commenced a recall of approximately ~140,000 (!) of their Bolt model electric vehicles after malfunctioning battery cells caused around 12 vehicle fires. It was later uncovered that the faulty cells were a function of one of two defects: a torn anode tab (the piece of the negative electrode that allows the cell to be wired into a module) or a folded separator (the thin sheet of material that separates the anode and cathode). In late 2021, the supplier of those cells, LG, agreed to reimburse GM for a whopping $2 billion in estimated costs and expenses associated with the recall. This was not a new phenomenon: in 2016, Samsung recalled roughly 2.5 million defective Galaxy Note phones, costing Samsung almost $5 billion, and in 2006, Dell recalled 4.1 million laptop batteries, costing the company approximately $400 million. The takeaway: the tolerance for error in electronics manufacturing is incredibly low, and the financial and reputational implications of a faulty product are incredibly high.

The Status Quo and the Promise of AI

One might think that these are one-off incidents and that manufacturers already deploy sophisticated automation systems that can mitigate such defects; however, this is seldom the case. Today, there are millions of human visual inspectors across global manufacturing supply chains (translating to billions in wage spend). These workers spend countless hours manipulating parts to manually identify defects. Not only is this approach prone to human error (according to OEMs we’ve interviewed, the best inspectors have accuracy rates of 85% - 90%), but it also presents a material health hazard. The need to focus on micron-sized product details under harsh lighting can cause significant eye strain and can ultimately lead to long-term vision damage. To combat these constraints amidst a tightening low-skill labor market, manufacturers have increasingly looked to technology-first solutions to solve their quality challenges. Among the most promising of these technologies is computer vision and applied AI.

In recognition of the magnitude of this problem and in alignment with our thesis to invest in novel technologies driving efficiencies in industrial contexts, the team here at SE Ventures conducted a thematic review of computer vision and applied AI for quality control. We actively assessed almost every major startup in this space over the past 24 months and ultimately identified UnitX, which we believe to be the technology and commercial leader in this ecosystem. Today, we are proud to announce our $5 million investment in the company, as part of their $46M Series B.

Why UnitX?

In our view, UnitX, through its software-defined vision system (“OptiX”) and its edge AI compute platform (“CorteX”), has developed the best AI-enabled quality inspection system in the market, both across incumbents and competing startups. OptiX?can configure dynamic lighting scenarios that improve image capture and illuminate potential defects more effectively than conventional lighting. Furthermore, the company’s CorteX product requires less training data (only ~5 images vs. alternatives that typically require 50+ images) to operationalize its deep learning models. These characteristics, in addition to a verticalized hardware + software approach, obviate the need for complex system integration across vendors, thereby decreasing the time to value for customers. UnitX’s superior performance and ROI have been validated by blue-chip customers that are highly discerning and have access to all the leading solutions in the market today. One customer notably suggested that UnitX helped improve the yield of their plant by 3%.

From a market perspective, we believe that the timing is perfect and UnitX stands to benefit from a massive global onshoring trend, particularly in high-cost labor markets like the US. While UnitX plans to serve every segment of discrete manufacturing in the future (automotive, semiconductors, aerospace, medical devices, consumer electronics, etc.), the company has achieved phenomenal product-market fit within its beachhead market: battery & electric vehicle OEMs. OEMs in these segments are benefitting from an influx of low-cost capital supported by policies like the Inflation Reduction Act (“IRA”). In the US alone, EV and battery manufacturers have collectively invested and promised to invest close to $100 billion in building domestic cell and module manufacturing. These companies promise to deliver an annual capacity of over 1,200 gigawatt-hours before 2030.

Most importantly, we backed UnitX because of the humility, patience, and drive of its stellar team, led by Co-Founders Keven, Max, and Adam. They have strong technical pedigrees that were shaped in the automotive heartland of Michigan and further refined with some of Silicon Valley's most notable technology companies. The three have managed to drive rapid product development while maintaining high-touch customer relationships; always prioritizing customer value above all else. We believe that the UnitX team is emblematic of our LP, Schneider Electric, through its customer-first DNA. We could not think of a better equipped team to tackle this challenge.

What comes next?

While UnitX was already well capitalized, the company recognized a compelling “better-together” vision with SE Ventures and Schneider Electric, which ultimately catalyzed the $5M investment from our fund. To drive tangible value beyond our investment, we see a significant opportunity to leverage our LP to help UnitX expand into new geographies and open new go-to-market channels. We are ready to activate our Operating Partner network within SE Ventures to help realize these synergies. As with all our investments, we are excited to embark on this multi-year journey with UnitX, and we are proud to be backing what we believe to be the next generational manufacturing technology company.

If you want to join the journey, UnitX is hiring — see here!