Other

Reflex Systems LLC

Subhash C Agrawal

Jonathan Hood Young

Scott M Wimer

Cash demand for a cash dispensing device is forecasted for individual days, weeks and multi-day load periods. In one embodiment, an expected value of withdrawal for any particular period, as well as an upper limit on expected withdrawals for a given confidence level, are computed. The technique facilitates the creation of a daily and multi-day withdrawal forecast for a given ATM. The forecasts are also used to compute an optimum load amount for a given load period (the period between loading operations) for a given ATM, subject to given parameters (such as courier cost, interest rate, transit time, desired confidence level, cost of run-out, and the like) and given constraints (bundle size, maximum capacity, possible delivery days, and the like). The forecasts are also used to determine an optimal load period length and schedule.

Method of and apparatus for forecasting cash demand and load schedules for money dispensers

A method of and apparatus for forecasting cash demand for an automated teller machine (ATM) or any other cash dispensing devices for individual days, weeks and multi-day load periods. In one embodiment, an expected value of withdrawal for any particular period, as well as an upper limit on expected withdrawals for a given confidence level, are computed. The invention facilitates the creation of a daily and multi-day withdrawal forecast for a given ATM. The forecasts are also used to compute an optimum load amount for a given load period (the period between loading operations) for a given ATM, subject to given parameters (such as courier cost, interest rate, transit time, desired confidence level, cost of run-out, and the like) and given constraints (bundle size, maximum capacity, possible delivery days, and the like). The forecasts are also used to determine an optimal load period length and schedule. After the ATM is loaded, the system monitors cash in the ATM during a load period to facilitate potential corrective action if actual withdrawals exceed forecasted withdrawals or, in the alternative, if the ATM will contain excessive funds on its upcoming regularly scheduled load date, e.g., due to much lower withdrawals than expected or due to the ATM becoming unavailable for a given portion of the load period.

A target software system is instrumented to generate behavior data representing a current observation or observation aggregate. A method then determines whether the current observation or observation aggregate warrants a second level examination; preferably, this determination is made by processing the current observation or observation aggregate through a first level detection algorithm that provides a provisional indication of a possible intrusion. If executing the first level detection algorithm indicates that the current observation or observation aggregate warrants a second level examination, the method continues by processing the current observation or observation aggregate through at least one second level detection algorithms to provide a more definite, fine grain indication of a possible intrusion. Multiple algorithms may be executed together within a single examination level, with the individual results then analyzed to obtain a composite result or output indicative of intrusive or anomalous behavior.

Growing community of inventors

Boxboro, MA, United States of America

Subhash C Agrawal