Coordinated by AIMEN, a research institute in Pontevedra in northwest Spain, the RETINA Project comprises a 13-member consortium of leading European researchers and technology companies. RETINA’s primary goal is to address a significant limitation in current sensor systems, namely that current sensor systems rely on either spectral imaging or lidar, which leads to incomplete data or a lack of versatility in varied environments. Accordingly, RETINA seeks to develop spectral imaging and lidar technologies and combine them together into a single system capable of providing a comprehensive analysis of both spatial (3D) and spectral (material composition) data in real time. RETINA aims to harness the power of photonic integrated circuits (PICs), quantum dots (QDs), and advanced CMOS and InGaAs detectors across the visible-to-near-infrared (VNIR) and short-wave infrared (SWIR) ranges.The aim is to couple these technologies with a digital infrastructure for machine learning algorithms to enable the development of multimodal sensing applications that offer unparalleled light sensitivity and spectral range. Novel PIC-based lidar uses tilted grating coupler arrays for non-mechanical beamsteering and FMCW modulation, fabricated on an InP substrate to provide the advantages of full integration in the chip, immunity to other lidar signals,and low power consumption.

The core idea is to fuse the two data streams (spectral and depth information) in real time through onboard signal-processing electronics and algorithms. By merging “what” (material and color properties from spectral data) and “where” (3D structure from lidar), the project seeks to create a versatile perception engine that can be embedded into applications such as:

• Autonomous Navigation: Enabling drones, robots, or self-driving vehicles to simultaneously recognize objects/materials (e.g., vegetation vs. pavement) and map their surroundings in 3D for safer path planning.
• Precision Agriculture: Monitoring crop health through hyperspectral signatures (e.g., water stress, nutrient levels) while mapping field topography to optimize irrigation or harvesting strategies.
• Environmental Surveillance: Detecting pollutants or hazardous materials (via characteristic spectral fingerprints) together with terrain profiling in remote or industrial settings.

Key development phases include:

1. Design and Fabrication (Months 1–18): Creating prototype spectral-imager chips (with integrated micro-filters) and photonic-chip lidar units (using silicon-nitride waveguides and modulators).
2. Algorithm and Firmware Integration (Months 19–30): Developing real-time data-fusion algorithms that align spectral frames with depth maps, plus on-device preprocessing to reduce data bandwidth.
3. System Integration and Packaging (Months 31–42): Combining both photonic chips, associated electronics, and optics into a robust, handheld or vehicle-mountable enclosure with thermal management and power supplies.
4. Field Trials and Validation (Months 43–48): Testing in real-world scenarios (e.g., greenhouse environments for agriculture, urban streets for navigation) to benchmark performance against separate commercial spectral imagers and lidar units.

The consortium includes academic research groups specializing in photonic-chip design, industrial partners with fabrication facilities for Silicon-Nitride (SiN) waveguides, and end-users from the automotive/agribusiness sectors. By the project’s end, the team expects to demonstrate a portable sensor head measuring under 10 × 10 × 5 cm, consuming less than 5 W, delivering real-time fused imaging at 10 Hz across a 50° field of view.

Read the full article in Photonics Views: https://onlinelibrary.wiley.com/doi/10.1002/phvs.202500006