Videos
Schneider Electric's Fahad Arshad on Why Data Readiness Decides AI at Scale
Recorded at the World Economic Forum’s Advanced Manufacturing and Supply Chains Forum 2026, this video was produced in collaboration with the World Economic Forum’s Lighthouse Operating System Academy.
Insights
- The main barrier to scaling AI in manufacturing is data readiness: clean, real-time data running on the right infrastructure to feed the models.
- Beyond productivity, AI enables risk-based cost avoidance, including predicting failures early and avoiding costly recalls that damage brand and customer loyalty.
- The biggest payoff is treating data as a strategic asset, so decisions on pricing and P&L happen in real time instead of waiting on long-term report analysis.
What turns a successful AI pilot into something that runs across the whole business? Fahad Arshad, who leads Global Industrial Digital Transformation in the SE Advisory Group at Schneider Electric, says it comes down to data. A model is only as useful as the information feeding it. Most early pilots run on data that is neither clean nor available in real time, so they stall before they scale. Get the underlying infrastructure right, Fahad argues, and the pilot finally has room to grow. He also makes the case that AI's payoff reaches well past productivity, provided the data foundation is there to support it.
Fahad Arshad:
My name is Fahad Arshad. I work for Schneider Electric, and I'm responsible for global industrial digital transformation within our SC Advisory Services.
Fahad Arshad:
Nowadays this is becoming increasingly challenging for a lot of manufacturers to get out of the pilot initial stage they started off from. And one of the most challenging parts which is resulting from this initial pilots is the data readiness. Yes, AI is really a fancy word and everybody wants to adopt it but the crux of it is really we need to ensure we have the right data infrastructure that enables us to have that clean data available, real-time data, to be fed into the AI and machine learning models which enable us to make those intelligent decisions and being able to scale those successful pilots at an accelerated pace.
Fahad Arshad:
Beyond productivity, there are three different layers we can look at. The first one is around the yield optimization, the overall waste reduction and overall throughput increment. Okay, so those are the things which help us really improve our overall volumes or overall production throughput, okay. So that's one being the one of the impact we can see overall business impact. Secondly we can look at what is really in terms of our risk-based cost avoidance. So many companies they're are able to now use advanced AI technology and those optimized machine learning models to predict early failures or early detection of anything happening which might be disruptive to their supply chain, their production, or even in some cases, being able to avoid the recalls, which can be very costly to a company in terms of the brand image, the loyalty, and recovering all those products from the market. So even though it's more of like soft savings, but being able to recall any such cost to the company and protecting the brand image for the company, that also can be quantified as a big investment saving. Third, I would say, the biggest thing is releasing data as a strategic asset. Because having the accessibility and being able to use data to make day-to-day real intelligent decisions, that goes even from making adjustments to their product pricing, looking at their P&L all these decisions, they can actually be more able to make them in real-time. And being able to use this as a benefit rather than having to rely on long-term analysis of reports and only then make some of those strategic decisions I would see beyond the productivity, these are some of the other quantifiable benefits which companies can really start to see when they adopt AI.
