Join Dave and Vlad with our guest today, Humera Malik of Canvass AI.

The conversation highlights the role of artificial intelligence (AI) in manufacturing, focusing on how the technology is evolving and its application in data-driven environments. Umaira Malik from Canvas AI shared her journey from the telecom sector to industrial AI, emphasizing the common challenge of data overload in manufacturing. Many companies have invested heavily in data infrastructure, collecting vast amounts of information from various sources like SCADA systems and data lakes. However, they often struggle to convert this data into actionable insights. The challenge isn’t just technical but also mental, as organizations need to shift their approach from simply hoarding data to leveraging it for decision-making.

One key takeaway from the discussion is the importance of contextualizing AI as a tool for augmenting existing processes rather than replacing them. AI's role in manufacturing often revolves around optimizing long-standing operations like fermentation or batch processes. In such cases, AI can be used to predict outcomes and improve efficiency without completely overhauling traditional systems. For example, Humera describes how AI was used to optimize a 40-year-old fermentation process by predicting when to end a batch earlier, allowing for better resource utilization without compromising quality. This approach emphasizes the practical, incremental benefits AI can bring to industrial operations.

The conversation also delves into the contrasting realities within the industry. While some manufacturing environments are advanced, equipped with process engineers who can handle sophisticated data analytics, many others are still in the early stages, using whiteboards or paper-based systems. These less digitized environments face challenges in understanding the potential of AI or even determining whether they have the necessary data to implement such solutions. The industry remains divided between those who are ready to adopt AI and those still grappling with foundational data issues.

Humera also discusses how generative AI (Gen AI) can play a role in accelerating AI adoption in manufacturing. By using Gen AI to analyze existing data and assess whether it’s sufficient to solve specific problems, companies can bypass lengthy data transformation processes. This allows them to quickly evaluate the feasibility of AI solutions and make more informed decisions. The shift towards leveraging Gen AI represents a significant advancement, helping companies move past the initial uncertainty surrounding their data capabilities and enabling more widespread AI adoption across different manufacturing environments.

Overall, the conversation reflects the growing importance of AI in industrial settings, the challenges of integrating it effectively, and the evolving tools that are helping bridge the gap between data collection and meaningful analysis. The focus on contextualizing AI as a supportive tool, rather than a disruptive force, aligns with the industry’s needs for gradual, value-driven improvements.

About Manufacturing Hub:

Manufacturing Hub Network is an educational show hosted by two longtime industrial practitioners Dave Griffith and Vladimir Romanov. Together they try to answer big questions in the industry while having fun conversations with other interesting people. Come join us weekly!

******
Connect with Us

• Humera Malik
• Vlad Romanov
• Dave Griffith
• Manufacturing Hub
• SolisPLC

#automation #manufacturing #robotics #ai