Team Name
Team JnJ Norderstedt/ Ethicon
Track
Manufacturing
Use Case
By leveraging production data feeding quality prediction models and an interactive optimizer dashboard, operators can continuously optimize assembly parameters, such as temperature and timing. This is key to ensure the highest standards of quality, reliability, and patient safety in every device produced. Ultimately, this digital transformation empowers Ethicon to meet soaring global demand for strategic medical devices, saving lives while maximizing operational excellence and business impact.
Technology
IoT Sensors Data Acquisition, Product Data Acquisition (SAS/ACCESS), Data Analytics and Machine Learning (SAS Studio, Visual Statistics, VDMML), Optimization (SAS Optimization), Dashboard and Visualization Tools (SAS Visual Analytics):
SAS technology powers the end-to-end optimization process by integrating production data analytics, advanced machine learning, and prescriptive modeling to continuously improve assembly parameters and reduce manufacturing waste. Interactive dashboards and optimization apps enable operators to simulate production scenarios, receive actionable recommendations, and update machine set points/controllable parameters, ensuring consistent quality and efficiency.
Region
EMEA
Team lead
@hschimke Schimke, Hannah
Team members
Schimke, Hannah; Dr Ismail, Aishah, Thys, Frederic (SAS)
Social media handles
Is your team interested in participating in an interview?
N
Optional: Expand on your technology expertise
Pictures and Videos on the Industrial Process had to be anonymized for confidentiality.
Why such a ML Pipeline:
Safety + credibility (GAM): Generalized Additive Models give transparent, monotonic effects of each variable and interaction between them ( pairwise surfaces). You can show operators and engineers why a recommendation works and enforce domain logic (Heatmaps were Optimal Machine Set points for Waste Reduction are identified by blue zones, with negative coefficients).
Hard constraints & guarantees (Optimizer): PROC OPTMODEL turns predictions into feasible, best-possible set points under engineering limits (bounds, ramps, rate-of-change, emissions, energy), with the ability to optimize multiple objectives (yield, energy, scrap).
... View more
