SBIR-STTR Award

The objectives of our Phase I effort are to characterize target sensor measurement uncertainties and feature extraction uncertainties; determine how and where, in the processing chain, these affect target discrimination and classification; show how the different sources of uncertainty lead to the cumulative uncertainty in the final decision; provide techniques that will be instrumental to optimizing the design of sensor architectures in order to minimize the effects of uncertainty; and demonstrate that our approach is an effective and efficient solution to determining what measurements and/or tracks should be exchanged between platforms in order to achieve a robust decision. Approved for Public Release 14-MDA-7979 (16 September14).

Keywords:
Robust Decision Making, Uncertainty Management, Multi-Sensor Data Fusion, Sensor Measurement Uncertainty, Classification

Our proposed second order uncertainty (SOU) product is a decision making software solution that addresses the problem of providing accurate and precisely defined decision courses of action (COAs) of complex, time-constrained problems in a fraction of the time required by alternative methods striving to achieve the same level of precision. Complex decision situations can deal with large volume of input data at near real-time speeds. Conventional techniques for achieving precision such a Monte-Carlo/Latin Cube simulations are limited in their ability to quantify the precision of the hypothesis in a timely manner. Our SOU product is the solution that will increase the confidence in the decisions at reduced decision times. There are no other methods in existence that provide both the accuracy and speed of SOU. Approved for Public Release, 15-MDA-8303 (1 July 15)