Chennai based maritime AI startup selected for global Smart Port accelerator in Singapore

CHENNAI: A city-based maritime AI startup, Dtyle.AI, has become the only startup from Tamil Nadu, and one of two from India, selected for the PIER71 Smart Port Challenge (SPC) 2026 Accelerate Programme in Singapore, a global maritime accelerator connecting startups with shipping companies and port authorities to pilot new technologies.

The IIT Madras-incubated startup was picked from a cohort of 22 startups across around 12 countries, including China, Denmark, South Korea and Singapore. The 10-week accelerator, running September 7 to November 12, is backed by the Maritime and Port Authority of Singapore (MPA) and NUS Enterprise.

Founded in February 2025, Dtyle.AI builds AI-powered vision systems for ships and ports, using computer vision and edge AI to detect early signs of fire, monitor crew safety and identify security threats in real time, unlike conventional systems that rely on sensors and alarms. “Being selected for PIER71 gives us an opportunity to work directly with global shipping companies and port authorities and validate our solutions through pilot projects,” Sasikumar told DT Next.

The team comprises marine engineers, chief engineers and AI specialists; its advisor is Professor Prabhu Rajagopal of IIT-Madras, a founding member of underwater robotics startup, Planys.

Sasikumar said that large container ships, car carriers and oil tankers typically run with just 20-25 crew, making round-the-clock manual monitoring of every deck virtually impossible. “Fire incidents on ships have been increasing, particularly with the rise in transportation of lithium-ion batteries and electric vehicles. Existing fire detection systems are reactive. Lithium battery fires escalate extremely fast, reaching nearly 1,000 degrees Celsius within seconds, leaving the crew barely five to six minutes to isolate the source before it becomes uncontrollable,” he added.

The system is built for maritime conditions of limited or no internet, low bandwidth and heavy radar interference. Instead of cloud-based processing, Dtyle.AI runs its AI models on onboard edge devices, letting them function independent of connectivity.

To cut down on extensive shipboard wiring, the company piloted wireless cameras integrated with its edge AI platform on an oil tanker at Chennai Port. “Our system detected smoke in 13 milliseconds, faster than the human eye. We also proved that it could function despite low bandwidth, radar interference and the unique operating conditions found on ships,” Sasikumar said.

For smart ports, it integrates feeds from thousands of surveillance cameras into a single dashboard, flagging only predefined events like unauthorised access to restricted areas, cutting the need for constant manual monitoring. It also targets stowaways, people who secretly board docked cargo ships, incidents that carry heavy financial costs since vessels may have to return them to their country of origin.

The platform can centrally track up to 100 ships from a single office, logging how quickly crew acknowledge safety alerts, data Sasikumar said could help shipping companies monitor fleet-wide safety compliance and potentially lower insurance premiums by demonstrating stronger risk management.

Each year, global shipping companies submit real-world operational challenges to PIER71, from which organisers shortlist problem statements for startups. This year, Dtyle.AI pitched solutions across three areas, autonomous vessel safety monitoring, privacy-preserving crew safety monitoring, and AI-powered smart port surveillance, securing its place in the accelerator.