CogniSense Themes

1 Sub-Terahertz MIMO Imaging: Explore energy-efficient MIMO radar at 140/220 GHz using CMOS/BiCMOS.
Deliverable: Dual-band LNA and PA prototype.
2 Multi-Wavelength Optical LiDAR:Develop adaptable solid-state FMCW LiDAR using integrated swept lasers and optical phased arrays.
Deliverable: Integrated LiDAR in GF 45CLO process.
3 Multimode Passive Imaging: Create ultra-low-power hyperspectral imagers across mmWave to IR.
Deliverable: Heterogeneous passive multi-spectral imaging platform.
4 Flexible RF/IF Circuits :Enable flexible RF processing to address high dynamic range and interference.
Deliverable: Silicon implementation of adaptive signal processing.
5 Converged Radar-LiDAR Array: Develop low-cost 3D-integrated radar-LiDAR front-end.
Deliverable: Packaged dual-mode pixel array prototype.

1 Sub-Terahertz MIMO Imaging: Explore energy-efficient MIMO radar at 140/220 GHz using CMOS/BiCMOS.
Deliverable: Dual-band LNA and PA prototype.
2 Multi-Wavelength Optical LiDAR:Develop adaptable solid-state FMCW LiDAR using integrated swept lasers and optical phased arrays.
Deliverable: Integrated LiDAR in GF 45CLO process.
3 Multimode Passive Imaging: Create ultra-low-power hyperspectral imagers across mmWave to IR.
Deliverable: Heterogeneous passive multi-spectral imaging platform.
4 Flexible RF/IF Circuits :Enable flexible RF processing to address high dynamic range and interference.
Deliverable: Silicon implementation of adaptive signal processing.
5 Converged Radar-LiDAR Array: Develop low-cost 3D-integrated radar-LiDAR front-end.
Deliverable: Packaged dual-mode pixel array prototype.

1 Hybrid Algorithms :Design algorithms for pre/post-digitization data reduction.
Deliverable: Software implementation.
2 Analog Feature Extraction: Develop circuits for real-time analog feature extraction.
Deliverable: AFE circuit blocks and test-chip.
3 Multi-Modal Fusion :Fuse AFE features to extract dynamics and reduce dimensionality.
Deliverable: Fusion algorithm emulations.
4 DFP Engine :Design low-power neural engines for real-time multi-modal feature streams.
Deliverable: DFP test-chip results.
5 Performance Analysis:Analyze A2I performance limits and system optimization.
Deliverable: Benchmark models and design guidelines.

1 Trust Estimation :Compute real-time trust levels for sensor data.
Deliverable: Algorithms and test-chip.
2 Adaptive Feature Extraction: Enable dynamic sensor adaptation to environment.
Deliverable: Software prototype.
3 Power-Trust Control:Design feedback systems for pixel-level power-trust scaling.
Deliverable: Simulation models.
4 Hardware Acceleration:Build efficient hardware for trust and adaptive control algorithms.
Deliverable: Accelerator test-chip results.
5 Power-Trust Management:Create circuits for optimizing sensing-performance vs. power.
Deliverable: Pixel tuning circuit measurements.

1 Power Management:Develop hierarchical, multi-domain power delivery systems.
Deliverable: Power management silicon results.
2 2.5D/3D Integration:Create 3D integration for compact, high-performance sensor systems.
Deliverable: Demonstrated 3D multi-spectral sensor.
3 ML-Based AMS Simulation:Use ML for large-scale AMS simulation accuracy.
Deliverable: Simulation algorithms and CogniSense results.

1 Sensor Control Framework: Develop host-processor-driven collaborative control software to optimize sensing-to-action operations.
Deliverable: Programming/control algorithms and simulation platform.
2 Sensor Reputation Tracking: Implement reputation scoring by integrating trust, domain knowledge, peer feedback, and past performance.
Deliverable: Software for reputation assessment.
3 Secure System Operation:Ensure robust system-level security against adversarial inputs and hardware/environmental threats.
Deliverable: Secure operation software stack.
4 System Prototyping :Build a drone-based demo platform showcasing CogniSense in real-time object tracking.
Deliverable: Functional prototype with enhanced perception/action capabilities.

About