Non‑Volatile Processing‑In‑Sensor Accelerator for AI‑Enabled Imaging Systems (A Non‑Volatile Processing‑In‑Sensor Accelerator for Imaging Systems, Tech ID: 23‑013)
Technology Overview: This technology integrates non‑volatile memory directly into image sensors, enabling processing‑in‑sensor (PIS) acceleration for AI‑driven imaging tasks. By performing computation at the sensor level, the system dramatically reduces data movement, latency, and energy consumption. The architecture is well suited for edge AI, smart cameras, and autonomous systems.
Industry Pain Point: Conventional imaging pipelines waste energy and time transferring raw data to external processors.
NJIT Solution: Processing‑in‑sensor architecture enables low‑latency, energy‑efficient AI computation at the edge.
Key Features & Advantages
- Non‑volatile in‑sensor computing
- Reduced data transfer and latency
- Energy‑efficient AI acceleration
- Ideal for edge and embedded vision systems
Development Stage: TRL 3–4 – Architecture‑level validation completed.
Target Markets
- Edge AI and smart cameras
- Autonomous systems
- Embedded vision hardware
Market Opportunity
- Global edge AI hardware market (2026): ~$15B
- CAGR: ~18–20%
- Projected market size (2035): ~$70–80B
Commercial & IP Details
Inventors: Shaahin Angizi, Arman Roohi