AI for a Safe Earth: Private Probabilistic Search and Predictive Threat Analytics

Abstract: The basic objective of a search is to identify an object and the position of the target. The target’s position may be uncertain or there may be complete or incomplete information about its location in terms of a probability distribution. The target may be stationary or in motion. The target distribution is associated with discrete or continuous search space. The problem of optimal search is to maximize the probability of detecting a target subject to the constraints of resources, effort and time. This work presents Private Probabilistic Search Mechanism (PPSM). The probabilistic search approach addresses the incomplete information on the target location by location probability. The problem is probabilistic from the perspectives of the location, size, distance and timing of the moving target(s) and distribution of the search efforts. The effectiveness of probabilistic search procedure can be verified on the basis of various properties of adaptive secure multiparty computation such as correctness, privacy, transparency, reliability and consistency. The search space can be divided into a set of private blocks; adequate number of sensors should be assigned to each private block; each block is monitored independently. This work highlights the complexity analysis of PPSM from the perspectives of computational cost and security intelligence. It also exercises case based reasoning on a test case of astronomical hazards and explores the scope of PPSM to assess and mitigate those threats. The universe is basically a computer, its history is being computed continuously. The astronomical hazards may be really dangerous threats against the sustainability of today’s human civilization and the existence of a safe earth. This type of probabilistic search problem is really hard to solve, it is not a trivial problem. It is also challenging to deploy PPSM in reality and seeks extensive support, coordination, planning and corporate social responsibilities from various space research organizations and earth science institutes globally. Artificial intelligence, Probabilistic Light Beam Search, Predictive threat analytics, Astronomical hazards, Reactive and proactive security, Private search, Adaptive secure multi-party computation