Alex Pamatat

IT Solutions / Head of Analytics COE , NXP

Supernova Award Category: 

Data to Decisions

The Organization: 

NXP Semiconductors N.V. (NASDAQ: NXPI), a global semiconductor manufacturer, creates solutions that enable secure connections and infrastructure for a smarter, better-connected world. NXP places a strong focus on research, development, and innovation, accelerating breakthroughs that anticipate, automate, and advance the markets they serve.  NXP provides 60 years of combined experience and expertise worldwide with locations in 30 countries. NXP's 29,000 employees serve automotive, industrial, mobile, smart home, smart city, and communications industries by providing semiconductor technology solutions.  In partnership with customers, NXP is not just advancing technology; they are advancing society.

The Problem: 

Prior to 2019, industry buzz words and “lingo” were the only data science currency exchanged at NXP; there existed no focused solutions to drive business value. Instead, IT’s business partners relied on legacy BI methods to solve business problems that became increasingly more complex. Such methods were often limited to univariate and bivariate statistics, limiting the depth of understanding and business value.

The Solution: 

What is the best way to apply advanced analytics across the organization? That was the question that ignited the experiment known as the "Citizen Data Scientist Program," a five-month strategic endeavor to up-skill existing data practitioners, helping cultivate the right analytics proficiencies. Achieving analytic proficiency required collaboration with senior leadership, the business community, and strategic vendor partners (Teradata, Dataiku) to improve citizen-led innovation through education, use case identification, coaching sessions, and a hackathon.  With leadership now witnessing the "tools" and" math" applied by their teams, and the potential value in the discovery and education process, leadership could correlate the connection between the "lingo" and the actual work completed or potential value delivered.

The Results: 

The CDS Program has been an ongoing initiative since 2019, with one iteration per year. NXP has been able to upskill 57 employees from 10 countries, applying 42 unique business use cases. A select subset of those employees converted to full or part-time data scientists. NXP's C-Suite leadership requested that IT leadership present a firm proposal in July to scale the program to serving 200+ employees/yr.  Now, participating organizations create their own embedded data science teams, with IT's Enterprise Advanced Analytics team acting as the de facto data scienceCOE. 

NXP's customers have also felt the impact.  Using customer quality as an example, before the CDS Program, many customers perceived NXP as a company out of fresh ideas and not at the desired competency to solve complex technical issues. As a result, customers had little confidence in NXP's abilities, in some cases, resulting in lost business and a descent down the customer quality ranks.  Being participants in each CDS program iteration, the Quality Organization used the data science approach and tools learned in the CDS program, which resulted in an immediate small win, further catalyzing additional investments in data science talent. In addition, the catalyzation resulted in the Quality Organization actively pursuing seven additional use cases where data science and advanced analytics are leveraged for business improvement and enhancing quality for customers.

Metrics: 

NXP started the CDS initiative with the objective of education but realized quickly the additional business benefit of the use cases developed. Utilizing the CDS program, the Quality Organization developed a use case to "Reduce the CI-Particles in Mold Compound," with results showing a significant business value increase through strong predictors of corrosion compared to years before implementing the CDS initiative. For example, before 2019, NXP was experiencing 45% corrosion due to CI particles. Utilizing a Machine Learning algorithm trained to identify strong predictors of corrosion, NXP reduced the 45% corrosion rate to 5%. Such substantial reduction in corrosion rates assisted NXP in eliminating high and medium-risk CI parts contributing to the long-term stability of products delivered to customers. NXP also achieved the goal to up-skill their data practitioners and educate the broader organization, demonstrating the business value of data science. Thus far, NXP has up-skilled 57 employees from 10 countries and has plans for an additional 200+ employees per year. This type of talent skill training will provide long-term substantial business impact and continues to be the catalyst for continuous business improvement.

The Technology: 

Using the standard definition of a technology stack NXP used primarily Teradata, Dataiku Data Science Studio, Hadoop, JMP, Python, R, Power BI, and Tableau. Building a solid data foundation in intelligent technology, such as Teradata, was a critical enabler. Equally important was providing technology for citizen data scientists. However, technology by itself was not the enabler; the process of integrating, training, and use cases with solid vendor engagement was the technology.

Disruptive Factor: 

For years NXP was experiencing a “cold start” problem. Unable to participate in data science due to no existing staff possessing the required skills and a leadership team that lacked understanding of what data science is or how it can be effectively applied and scaled across a large enterprise.  NXP had concerns about investing in a concept and tooling that the existing talent resources and leadership did not understand.  The Citizen Data Science Program produced the much-needed evidence and examples of data science, using real-world NXP use cases, with a low financial barrier to entry due to shared program resources.  With senior leaders now realizing the value of data science, they can rationalize the required additional investments to scale data science efforts within the organization and now have a base competency to build upon.

Shining Moment: 

Many CDS participants were propelled into new career paths, benefiting NXP and their personal development. Many continued their training beyond the program and transitioned into functional data science roles. The program provided org awareness & personalized business value from data science to sr leadership and exec teams. The CDS program made DS approachable, enabling NXP to execute DS on real business opportunities, enabling a catalyst for a collective analytics awakening.

About NXP

 

NXP Semiconductors N.V. a global semiconductor manufacturer, creates solutions that enable secure connections and infrastructure for a smarter, better-connected world. As the world leader in secure connectivity solutions for embedded applications, NXP is driving innovation in the automotive, industrial & IoT, mobile, and communication infrastructure markets