Dong Hyun Lee, Korea Regional Sales Director, proteanTecs

At Automotive World 2024, according to Vijaykishan N, Vice President at proteanTecs, their solution represents a healthcare system for 'computers on wheels' and extends to all electronic devices. Following the exhibition, I had a more detailed conversation through Dong Hyun Lee, Korea Regional Sales Director. proteanTecs' on-chip monitors strategically placed during chip design provide high-resolution images of chip operation, performance, and status. This generates purposeful deep data with machine learning algorithms and offers crucial insights for fault prediction and maintenance through cloud and edge analytics solutions.

WRITTEN BY Sang Min Han

한글로보기







Could you please briefly introduce the company along with its major milestones? 
Lee          proteanTecs is the leading provider of deep data analytics for advanced electronics monitoring. We provide system health and performance monitoring, from production to the field. By applying machine learning to novel data created by on-chip monitors, our analytics software solutions deliver unparalleled visibility and actionable insights - leading to new levels of performance, quality and reliability.  

Founded in 2017, proteanTecs emerged from stealth mode in 2019. The company is headquartered in Israel and has offices in the United States, India, Taiwan, and most recently, South Korea. proteanTecs has raised over $200 million from world-leading investors, has nearly 250 employees, and is trusted by industry-leading customers in the datacenter, automotive, communications and mobile markets.



What I understood from the proteanTecs booth at Automotive World 2024 is that its solution revolves around the monitoring of chips and systems, which are central to both current and future automobiles. It involves deep data analysis for checking the health and performance of automotive chips, including chips, algorithm, and cloud solutions. Is that correct? Automobiles already have various software and hardware standards for safety, including ISO 26262. What are the limitations of these standards, and why is proteanTecs' solution crucial? 
Lee          The automotive industry is currently experiencing a transformation driven by electrification, connectivity, and autonomous driving. Vehicles require extensive computational capabilities and generate massive amounts of data. Future cars will embrace new architectures, becoming software defined computer on wheels capable of hㅇsting advanced applications, machine-learning algorithms and continuous reconfiguration. Ongoing advancements in software and hardware, as well as their interface, are pushing the boundaries of performance and functionality. However, the same technology that will bring this vision to life also introduces new challenges.

Unfortunately, conventional predictive and preventive maintenance approaches fail to scale down into advanced devices to meet the safety, performance and cost challenges faced by vehicle manufacturers, OEMs, and fleet managers. Current methods rely on production testing to ensure lifetime performance, but that doesn’t follow the system into the field, where environmental and operational conditions take a toll. Thus effects of device aging, application stress, latent defects and other lifetime events cannot be detected. In a market where safety is key, hardware is expected to function for many years, but software frequently changes, maintaining reliable and prolonged mission profiles becomes even more challenging.

In response, proteanTecs developed solutions that not only enable predictive and preventive maintenance, from within, but transform automotive electronics into intelligent system sensors. The company is enabling in-situ monitoring of the device while in mission-mode so as not to disrupt normal operation, but also to capture all phenomena under real functional workloads. This is the only way to ensure the road to zero failures.

Automotive safety standards, including ISO 26262, are now starting to include predictive maintenance compliance in the standard. proteanTecs monitors have received ISO26262 Automotive Safety Integrity Level B (ASIL-B) certification, reflecting our ongoing commitment to supporting mission-critical systems at the highest standards. 










How does proteanTecs' solution analyze and predict the deep data from chips to automobiles and the cloud? Where does its core technology lie? 
Lee          proteanTecs on-chip monitors - known as Agents - are strategically placed during chip design to provide a high-resolution picture of a chip’s behavior, performance and health. These Agents generate purposeful deep data to which machine-learning algorithms are applied and made accessible in proteanTecs’ cloud and edge analytics software solutions. 

The company provides a suite of applications for in-car safety monitoring and for cloud fleet-level analysis. Their embedded on-board and in-chip applications provide real-time, low latency health and performance indications, while the cloud platform provides big data analytics for predictive maintenance.

Users gain actionable insights and alerts around the chip and system’s performance, quality and reliability, in both test and mission modes. This capability extends across the full lifecycle - from system-on-chip (SoC) characterization and high-volume manufacturing to system ramp and production, and in-field operation. 




Is it not just about predicting chip failures? How does managing chip health, handling errors, and providing various derived benefits contribute to automotive customers? 
Lee          The key to preventing failures (the inability to perform properly) is to identify and resolve faults (abnormal behavior or defects) before they become failures. It’s important to note that, diagnostic practices detect faults only after they’ve become permanent, and by then failure may be just seconds away. Typical descriptive and diagnostic analytic systems can’t always pinpoint the root cause of a problem, which leads to time-consuming and costly trial-and-error processes.

In contrast, advanced predictive maintenance methods predict what will happen in the future and present the best way to prevent problems from happening. They leverage sophisticated monitors that implement many data points as high-quality performance indicators. Analyzing that data can reveal when a system is in a critical state. Alerts are generated days, weeks, or even months before an impending failure happens, allowing users time to take remedial action before a fault becomes permanent, and before a failure ever occurs.

Automobile OEMs must take every possible measure to guarantee reliability and safety, yet the sheer number of electronics brings an increasingly high risk of system failures. When failures do occur in the field, it’s crucial to identify the exact cause of problems, whether it’s defective hardware or software components, or other issues. proteanTecs’ continuous performance monitoring, mission profile monitoring and degradation monitoring solutions make it possible to generate, extract and analyze data from SoCs at the heart of today’s most advanced automotive electronic systems. 

By using proteanTecs’ technology, automotive companies can: 

· Gain greater insight into the health of electronic control units (ECUs)  
· Monitor performance degradation and predict time-to-failure   
· Pinpoint the source of faults 









Developing such a solution requires understanding various fields like semiconductor manufacturing, system implementation, software, and data analysis. How is all of this possible? Where did the expertise and legendary leaders come from? 
Lee          Our founders are technical veterans of the electronics industry and serial entrepreneurs. Their previous company, Mellanox was acquired by NVIDIA for $6.9 billion. They witnessed firsthand the rising challenge of achieving quality and reliability without affecting cost and performance, especially as technologies advanced, and knew there was a better way. They founded proteanTecs with a mission to enable the electronics industry to continue to scale and compiled a team of leading multidisciplinary experts in the fields of chip design, machine learning and analytics software to transform their vision into reality. 



What form does the business model take? 
Lee          proteanTecs offers several software solutions and the company charges a subscription fee for the various applications. There is also an integration fee for embedding the technology into the system.



The customer base and partners of proteanTecs, such as OEMs, Tier 1 companies, automotive software companies, and semiconductor companies, seem to be very diverse. Could you please introduce which companies you are working with and what you are doing with them? 
Lee          In terms of the automotive industry, multiple automotive chip manufacturers and OEMs are using proteanTecs to enhance their chips with continuous reliability and performance monitoring. Leading automotive OEMs are also pushing down these requirements into their supply chain as they move toward the goal of zero downtime. 

However, our deep data analytics solutions can benefit many end applications. From big hyperscalers to mobile manufacturers, aerospace & defense to telecommunications, there are many issues involving variability, downtime and safety related to electronics failures and cost efficiencies, and proteanTecs directly addresses these challenges. It is increasingly difficult to develop a chip or a system using the most advanced process technologies and packaging, while meeting all the requirements of mega-functionality, field reliability and power/performance.

Our comprehensive monitoring solutions require ecosystem collaboration to drive and accelerate market adoption, which is why we have partnered with several industry leaders across the value chain. Our partner page on our website lists these partnerships. We’re fortunate to count industry-leading companies, like HARMAN Automotive, TTTech Auto, and Dell Technologies, as our partners. These collaborations further our commitment to helping the end customer get the most visibility and insights out of their product.



It seems that you may already have connections with Korean companies. If there's anything you'd like to convey to Korean companies, please do so. 
Lee          The Korean electronics market is a great fit for our technology. We recently opened a proteanTecs office near Seoul and are strategically expanding our local team to support our customers. We are actively working with multiple Korean companies, ranging from innovative startups, like FuriosaAI, to the top-tier Korean electronics brands. In the case of FuriosaAI, they selected proteanTecs’ deep data solutions to ensure the performance and field reliability of their cutting-edge inference chips and are using our technology in their next-generation AI accelerators targeted at hyperscale data centers. 

I encourage Korean companies to contact our team if they are interested in learning more about proteanTecs’ solutions. 




At Automotive World 2024, according to Vijaykishan N, Vice President at proteanTecs, their solution represents a healthcare system for 'computers on wheels' and extends to all electronic devices.

AEM_Automotive Electronics Magazine