Japanese IT firm NTT develops AI chip for drone video analysis

Nippon Telegraph and Telephone Corp. (NTT) has developed an AI chip that processes high-resolution video captured by drones.

The chip enables real-time, low-power analysis and can detect objects or individuals from heights of up to 150 meters.

This technology is aimed at applications like infrastructure inspections, neighborhood patrols, and live event coverage.

NTT plans to make the chip available for commercial sale by March 2026.

The company has previously invested in photonic chips and optical AI processors. This shows NTT’s focus on advancing AI technologies.

🔗 Source: Bloomberg

🧠 Food for thought

1️⃣ NTT’s entry builds upon four decades of specialized computing experience

NTT’s move into AI chips isn’t a sudden pivot but builds on deep institutional expertise in specialized computing hardware.

The company developed ELIS, a LISP machine specifically designed for AI applications, as far back as 1985, making it an early pioneer in purpose-built AI computing hardware 1.

This AI chip announcement also follows NTT’s significant work in photonic technology, including their development of compact, low-power optical modulators that demonstrated speeds of up to 32 Gbit/s while maintaining low insertion loss 2.

NTT’s innovation approach typically combines both hardware and software development, as shown in their historical projects that integrate specialized architecture with application-specific capabilities.

The company’s experience with semiconductor device reliability, dating back to transistor implementation in telecommunications equipment in the 1960s, provides crucial expertise for developing robust AI chips 3.

2️⃣ Drone-focused AI processing targets a rapidly expanding market segment

NTT’s focus on drone video processing taps into a market projected to grow substantially, with the global drone industry expected to reach $244.95 billion by 2032, growing at a CAGR of 22.4% 4.

AI integration in drone technology specifically is forecasted to become an $84 billion market by 2030, driven by applications in emergency services, infrastructure inspection, and precision agriculture 5.