The paper is devoted to the optimization of counteraction to an object evading detection by stationary sensors in a threat environment. It is assumed that the evading object moves optimally between two given points. Placement of sensors in a given rectangle between start and finish points is required so as to maximize the minimum possible value of the risk functional for the evasive object. The problem is analyzed using a combination of optimal control and gradient methods. The results of numerical simulations for the case of arranging three sensors in different rectangles are presented.