How Data Sharing and AI Could Reduce Downtime in Smart Factories


Confirm researcher Dr. Sourabh Bharti wants to create collaborative ecosystems for data in smart manufacturing, which could provide insights on quality and efficiency and help predict downtime.

After receiving his Bachelor’s and Master’s degrees in computer science and engineering, Dr. Sourabh Bharti completed a PhD in information technology at the Indian Institute of Information Technology and Management in Gwalior. During his doctoral studies, he also worked as a visiting researcher at Anglia Ruskin University in the UK and received a fellowship to undertake partial doctoral studies in Hungary.

He then received a Marie Sklodowska-Curie Actions Fellowship in 2020 and joined Confirm, Science Foundation Ireland’s research center for smart manufacturing. Currently, he works at the Nimbus Research Center for Cyber-Physical Systems and IoT at Munster University of Technology in Cork.

“Our investigation is an attempt to encourage organizations to engage in a secure intelligence sharing exercise without compromising the privacy of their business data.”

What inspired you to become a researcher?

It was in 2012 when I was working on my master’s thesis. My supervisor introduced me to the world of research, which gave me the freedom to explore and express independent research ideas.

I became familiar with posts and authorship, which was a perfect gateway to connect with the outside world. Also, being a researcher does not allow me to settle down and that is why every day there is something new to learn and explore, which is what I like the most about research.

Can you tell us about the research you are currently working on?

The project focuses on performing distributed machine learning on edge devices deployed in the smart manufacturing space to collect various parameters related to industrial assets and operations.

The idea for this project arises from industrial data silos that remain spread out across locations and are difficult to gather due to industrial competition and data privacy concerns. This pushes organizations to process data closer to its origin (at local manufacturing sites), but still a single manufacturing site cannot gather all the necessary standards for applications such as asset failure prediction and evaluation. of quality.

This creates a collaborative ecosystem where multiple participants belonging to different manufacturing sites in the same organization or from different organizations agree to share their intelligence without revealing their raw manufacturing data. This is currently being accomplished through distributed machine learning techniques, such as federated learning.

Our project is focused on making federated learning suitable for resource-limited edge devices by employing lightweight predictive modeling and enabling resource-conscious computational offloading and preserving privacy when required.

The project is in line with the EU data strategy, which advocates a shift from centralized cloud-based processing to edge-based processing. We are working closely with our industry partner IBM to gather information on current industry initiatives being taken to drive edge computing. We observe and work on some fascinating edge computing use cases, such as a mobile product quality inspector developed by IBM in collaboration with Apple.

The project may offer some cutting edge solutions that the industry can adopt right away! It could also increase the visibility of Confirm as it is still an emerging area and not much effort has been made in the past to investigate collaborative ecosystems like this for smart manufacturing.

In your opinion, why is your research important?

Traditionally, manufacturing organizations process their data in silos. Reasons include increasing industrial competition and a reluctance to share sensitive manufacturing data.

Our research is an attempt to encourage these organizations to engage in a secure intelligence sharing exercise without compromising the privacy of your business data. The incentives can come from the reduced manufacturing cost generated by improved intelligence and incentives at the organization level.

What commercial applications do you envision for your research?

There are numerous business aspects of a collaborative ecosystem in manufacturing, as shared intelligence can belong to different domains, such as asset failures, energy efficient manufacturing practices, and so on. their occurrence, which can reduce unplanned downtime.

From a product standpoint, edge-based smart manufacturing products such as mobile quality inspectors are in high demand as they eliminate the reliance on a subject matter expert to inspect product quality. Furthermore, they can easily participate in a collaborative learning process without revealing their raw data, as is currently being done in other applications, such as Google’s text-based predictions.

What are some of the biggest challenges you face as a researcher in your field?

As the collaborative ecosystem is in its infancy, organizations are reluctant to move away from traditional data processing practices. The manufacturing industry has yet to embrace such collaborative ecosystems as it requires processing manufacturing data closer to its source, which is not currently practiced.

From a research and academic development point of view, the biggest challenge is the lack of relevant data sets to test the designed approaches.

What are some of the areas that you would like to see addressed in the coming years?

I hope that the manufacturing industry will fully adopt edge-based data processing to get real-time pattern mining.

This is going to be a big change from centralized data processing, but EU initiatives like the European Data Strategy really put cutting edge computing and collaborative data spaces into perspective.

Don’t miss out on the knowledge you need to be successful. Sign up for the daily letter, Silicon Republic’s roundup of necessary science and technology news.


Please enter your comment!
Please enter your name here