ai
Discover what separates AI success from AI hype. In this series from MIT Sloan Management Review, AI winners share their secrets and success stories from the front lines. Explore the future of artificial intelligence with leaders from companies like YouTube, Cisco, and Hugging Face who are turning AI's potential into measurable business value.
Suresh Venkatarayalu shared that Honeywell's Forge platform has scaled up significantly, connecting 5.3 million assets as of July 2026 compared to 700,000 assets two years prior. He noted that the company is working to connect its estimated 11 to 12 million total install base faster using new ontology approaches.
Suresh Venkatarayalu explained that Honeywell is layering probabilistic artificial general intelligence models on top of its traditional 140-year-old deterministic and heuristic control systems. This combination allows the company to build learning systems that still adhere to mission-critical safety standards.
Suresh Venkatarayalu discussed the creation of Maintenance Assist, an AI tool trained on legacy service tickets and technical manuals to help field technicians. He stated that the system started at roughly 50% accuracy and improved toward 85%, resulting in a massive productivity uplift.
Suresh Venkatarayalu detailed how Honeywell incorporates AI intelligence into industrial plants, such as a major ethylene plant processing thousands of sensor data points per second. The system cuts 20% to 60% of false positives from sensor inputs to guide operators more accurately.
Suresh Venkatarayalu noted that connecting a million-square-foot building used to take 120 to 150 hours two years ago. Through the deployment of Forge Academy and specialized training for system integrators, connection times have been reduced to between 10 and 20 hours.