Can AI Make Physical Operations Smarter? (ft. LTA)
Join us for a special AI ToolsDays series exploring how embodied AI, computer vision, and edge intelligence are reshaping physical-world engineering. Hear from practitioners building AI tools for mechanical design, manufacturing, and deployment on constrained hardware.
More About the Sharings
Chang Yao (Founder, Buildables) will share on "Building AI for Mechanical Engineering Workflows"
While frontier AI models have transformed software development, applying them to mechanical engineering presents a very different set of challenges. Chang Yao will share more on Buildables' AI copilot for industrial hardware and demonstrate how domain-specific AI models and engineering workflows can support design review, technical knowledge retrieval, and manufacturing optimisation, helping engineers move more efficiently from concept to production.
Dr. Kan Shyi Herng (Managing Director, Helios Applied Systems) will share on "AI-Generated Vision Models for Industrial Inspection"
Configuring industrial machine vision systems traditionally requires extensive tuning of camera parameters, lighting conditions, and image processing pipelines, often taking hundreds of engineering hours to achieve reliable performance. Shyi Herng will demonstrate how AI can automate this process by generating embedded vision models directly from uploaded images. The system evaluates more than 64,000 possible operating conditions, accounting for camera characteristics, environmental variations, and difficult-to-decipher alphanumeric codes, before producing an optimised model capable of robust, high-accuracy code reading in production.
Vinod Rajendran (Generative AI Expert, AUMOVIO) will share on "Efficient AI on Edge Devices"
Automotive companies are increasingly demanding cost-competitive, AI-enabled high-performance compute (HPC) devices. Such devices typically operate under tight compute and memory constraints, making the development of efficient AI technologies critically important. Hear how AUMOVIO have developed tools and systems for AI model optimization and compression to run efficiently on resource-constrained hardware. This includes a perception system and a conversational AI assistant powered by a small language model — both designed to run on automotive-grade System-on-Chips (SoCs).
Osman Ahmed (Co-Foudner, Curium) will showcase "From Reactive Monitoring to Predictive Flow."
How can operators anticipate congestion before it becomes visible? Osman will share how Curium combines LiDAR with existing CCTV to build a privacy-preserving digital twin of busy public spaces, enabling real-time crowd flow analytics, congestion prediction, and safety monitoring without relying on facial recognition. Through deployments across transit hubs and large public venues, he'll discuss the design considerations behind building AI systems that can support operational decision-making in the physical world.
Jia Yi (Founder, Embodied Solutions) will share on "Delivering Embodied AI at Scale."
While advances in embodied AI have made robots increasingly capable, deploying them reliably in production remains a significant challenge. Jia Yi will share how Embodied Solutions bridges the gap between manual operations and full autonomy through a platform that combines teleoperation, data collection, AI training, and fleet management. Learn how this approach enables organisations to deploy and continuously improve AI-powered robotics without having to build the underlying infrastructure from scratch.
Siddharth Krishnan (Founder, The Robot Company) will share on "Deploying Embodied AI in the Real World."
Sidd will demonstrate their teleoperation platform for humanoid robots. Rather than focusing on a single application, the platform is designed to enable robots to perform a wide range of real-world tasks. Through videos of the same robot tackling everything from household chores to data centre cabling, alongside footage from enterprise deployments, Sidd will explore what general-purpose embodied AI looks like in practice and why versatility matters when deploying robots at scale. Also, catch a live demonstration of one of their humanoid robots!
*Do note that content and flow might be subject to changes.
Programme
2.30pm - 3.30pm: AI solution demos and showcases
products/research in Computer Vision, Embodied AI, and related technologies
3.30pm - 4pm: Sharing of LTA problem statements
4pm - 5pm: Pitch to Problem Owners & Networking
Meet the LTA teams behind each challenge area and share how your solution could help address their operational needs. Solution providers are encouraged to visit the booth(s) most relevant to their work.
"Lights-Out" Depot: AI-powered inspection & diagnostics, Automated materials handling, Workflow orchestration
Self-Monitoring Mainline: Infrastructure inspection, Defect detection, Predictive maintenance, Dynamic work assignment
Intelligent Station Operations: Crowd management, Safety, passenger experience, operational decision support
More About the Speakers
Chang Yao is the founder of Buildables that makes mechanical engineering models for critical sector companies. Their latest model, Orion, is being deployed for operational and design tasks in Singapore and East Asia.
Dr. Kan Shyi Herng is a distinguished technologist and the architect behind Helios’ core innovations, including machine spatial awareness, industrial process AI agents, and autonomous robotic systems. He holds a PhD from Cornell University have extensive expertise in advanced manufacturing, leading successful digitization and automation initiatives across the semiconductor and aerospace sectors for industry leaders. Beyond his industrial achievements, Dr. Kan serves as an Industrial Advisor at the National University of Singapore for the Engineering Science (ESP) and Master of Science in Precision Engineering (MPT) programs. A dedicated mentor to students and engineers alike, he is committed to shaping the next generation of talent and the future of AI-driven manufacturing.
Vinod is an experienced AI specialist who drives innovation projects in the area of Generative AI tailored for next-generation automotive applications. Prior to AUMOVIO, he was a Machine Learning / Data Scientist at Siemens Mobility and Rolls-Royce. He holds a Master's degree in Computer Science from the Technical University of Munich (TUM), Germany, with a focus on Artificial Intelligence.
Osman Ahmed is the Co-Founder and CEO of Curium, where he leads the development of AI-powered sensing and perception systems for real-world environments. His work focuses on privacy-preserving spatial intelligence, helping organisations turn live sensor data into actionable operational insights across transport, public spaces, and smart infrastructure
Jia Yi has a bachelor's in mechatronics and master's computing. In 2015 at age 16, he co-founded the algorithmic trading startup KloudTrader and in 2023, he created the note generation app FutureNotes. He also launched the open-source adult-size humanoid robot project OpaNoid in 2020, which continues to serve as a key motivation for Embodied Solutions.
Siddharth Krishnan is is the co-founder and CEO of The Robot Company (TRC), a forward-deployed robotics startup based out of the US and Singapore. TRC focuses on bridging the deployment gap in robotics through a platform that lets enterprises put robots to work today, while building the data engine for autonomy tomorrow. Before TRC, he invested in and built ventures for Temasek.
More About the Series
Curious about the growing world of AI tools and how they’re shaping different industries? AI ToolsDays is our beginner-friendly, demo-focused series organized around specific themes, showcasing AI tools in action through practical walkthroughs and real-world use cases.
Get involved: Learn more about Lorong AI | Speaker Sign-up | WhatsApp Community | LinkedIn | X