Industrial artificial intelligence is moving beyond software assistance into real-time engineering and machine control systems, even as large-scale deployment of agentic AI in engineering remains at an early stage, Ashish Khushu, Chief Technology Officer at L&T Technology Services, told TNIE.
He was speaking on the sidelines of the company’s ninth edition of TECHgium, an engineering innovation platform for students. The latest edition recorded over 62,000 registrations from 540-plus engineering institutes nationwide, marking a 60% surge over the previous edition and nearly 800% growth since its launch in 2016.
Khushu said engineering-focused AI systems differ significantly from software-centric AI tools that dominate most enterprise conversations today.
Khushu described one industrial safety project involving a wood-cutting machine used in the US market, where AI is used to prevent accidents during operation.
The system uses a camera mounted above the blade to monitor hand movement near the cutting area. If the operator’s hand enters a defined danger zone, the system detects the motion, processes the image and sends a command to stop and retract the blade.
“All of this happening within 18 to 31 milliseconds,” Khushu said, describing the project as an example of AI working with imaging systems, embedded chips, electrical instrumentation and mechanical systems simultaneously.
He said such applications highlight how AI is increasingly being used in engineering environments where systems must respond in real time rather than simply assist with software tasks.
Khushu also said engineering companies are using AI to shorten product development timelines and improve product quality.
“One area where AI is really helping and not very often taught is the initial part,” he said, referring to product design specifications and engineering requirements gathering.
“If you can capture the functionality, the definition of these products that you are designing very acutely in minute details and there is a lot of accuracy, you really are solving a huge problem of development,” he added.
He said engineering firms are increasingly looking at reducing product design cycles from “three years to 18 months” or “five years to two years” using AI-led interventions across development stages.
TECHgium 2026
Finalists at TECHgium this year presented projects including robotic systems for medical diagnostics, AI-based multilingual video translation, live captioning tools, Wi-Fi sensing for smart home monitoring, and robotic arms designed to improve industrial safety. After a nine-month mentoring and evaluation process involving LTTS engineers and industry experts, 34 teams presented working prototypes before a jury of industry leaders, academics and analysts. Winners received prizes worth more than Rs 18 lakh.
A team from M. S. Ramaiah Institute of Engineering won for its project on a self-charging hybrid scooter. Students from RNS Institute of Technology finished runner-up for “TETROBOT”, a modular robotic arm, while PSNA College of Engineering and Technology secured second runner-up for “AERIS”.