Session 3 – 27 May

Next stop, R&D is digital – AI is the tool to digitize R&D

16:15 – 16:55

R&D organizations are continuously pushed for efficiency, shorter innovation cycles and more output. Hyper automation is the way to drive out costs and speed up development cycles. Using automation is for a long time the de facto standard for removing costs in business processes and lowering the overall bottom line. Typically, we see this being applied by the IT department in what they call automation of white spots.

Inside R&D organizations, we often see very bright people automating all our customers’ needs but not so much their own way of working. That’s about to change. But purely automating manual or error-prone processes won’t completely address the needs in an R&D organization. Knowledge management, or rather the lack of knowledge, is one of the major sources of waste, delays and being out of control, creating a liability towards the future. The ultimate goal is to automate and capture knowledge at the same time – knowledge that can be used far into the future. We believe using AI as an automation tool will address both needs: efficiency and knowledge.

In R&D, we see knowledge as the major asset to protect and nurture, especially in large systems with many millions of lines of code and long lifetimes (> 25 years). One of the processes hampered by knowledge erosion is defect handling: triaging, analyzing and solving defects. Capturing knowledge during the early stages of product development in a model, using artificial intelligence, will help us drive efficiency now and in the long run. The knowledge we capture now will be our ‘consultant/mentor’ in many years from now when we still need to maintain and fix issues on what has become a legacy system.

Together with ESI (TNO), Philips has developed an approach for automated defect handling. In the short term, it will help us by driving efficiency in wrong assignments, duplicates, rejects and even linking defects to faulty code based on defect descriptions. Towards the future, the model will be our expert to analyze and solve defects.

Gernot Eggen is currently heading the R&D organization within the imaging chain cluster at Philips Image-Guided Therapy. He’s also responsible for software engineering at IGT. The two roles are an ideal combination in which his passion for innovating and (disruptive) software technologies flourishes. Since the early days, he’s been an advocate for software modeling, model-based testing and artifact generation. In 2019, he started to apply AI in R&D settings.

Patrick Bronneberg is a tech-savvy department manager. In the past 14 years, he’s worked at Philips in roles from software engineer to software architect and department manager. In his career, he’s contributed to embedded software for imaging devices, algorithm development for in-vitro diagnostics, cloud platforms for population health, IoT for personal health and now machine learning. As a department manager, he gets energized by initiating improvements – typically, by focusing on technical excellence and adoption of the newest technologies.

Gernot Eggen 

Senior director imaging chain cluster at Philips IGT Systems

Patrick Bronneberg

Technical department manager display software at Philips IGT Systems