Demand-filling of ATMs

Cost savings by automating cash cycle processes

CineoThe total cost of cash handling in Germany amounts to several billion euros a year. Cost savings can be made in this area if the cash cycle process can be automated. Cash handling relates to various economic sectors. This article relates to cash handling in connection with ATMs in the non-banking sector.
It is important that these ATMs are optimally filled for economic reasons. Thus, costs can be saved and additional interest income can be secured. However, it has been observed that decisions regarding the filling and emptying of ATMs are often made on the basis of experience and gut feeling. If the key people are not available, it is very hard to find replacements because the necessary know-how and experiences are not usually to be found in other people.
In demand-optimized filling of ATMs, maximum cash availability must be ensured on the one side, so that there is always sufficient withdrawable money available to satisfy customer requirements. On the other hand, related costs should be kept as low as possible.
Costs arise from the filling of ATMs because of high money holdings in the machine. These stocks belong to the tied-up capital of a company, which result in a loss of interest income. Costs also arise from the transport of valuables. Therefore, transport should be kept to a minimum.
A concept for application software is introduced below in which an existing application for ATM filling is optimized. The concept should empower the administrators of ATMs to determine their own needs as precisely as possible and then accordingly plan the filling of the machine. The application should enable the demand for money for individual days and weeks to be calculated as easily and therefore as accurately as possible. By using the demand amount identified, determining the order date, order quantity and the delivery date can be automated.

Demand assessment with the help of different categories

BedarfsermittlungTo fill the ATMs as required, initially the future demand for money has to be determined. Different approaches were used to assess demand. Determination by means of various daily and weekly categories turned out to be the most suitable.
In this regard individual days and also weeks were defined. Demand is always set to a category.
Days categories can be specified, for example, as regular business days, weekend days or holidays. The week categories are similar, they can be defined as a normal week, for example, or a week in the run up to Christmas, or weeks during holidays.
An arbitrary number of categories is fixed and then these are assigned to individual days or weeks. A one-week class is stored as a standard, so that in times when the demand has not been directly associated, a demand can always be assigned. This is important for further calculations.
It is clear that the concept presented in this variant is very flexible and can adequately cover exceptional situations. At the same time it is also clear that requirements can be ascertained quite accurately with little expenditure of time.

Determining order date

To fill an ATM on demand, the optimal order date must be identified. Specifically, you need to determine on which days it is economically effective for the users of the application to place an order.
To determine this day, data on the order lead time of the secure carrier (SC), working days on which an order can be placed and weekdays on which it can be delivered must be integrated. The data on these conditions can be stored in a configuration file.
To determine the economically most effective day for a cash order, first assume that it is most effective to order on the last possible day. This results from the fact that as little capital as possible should be bound up in the provision of sufficient money to meet demand. It is also advantageous to calculate the order amount as accurately as possible.
The order date determination process is illustrated again with an example.
In a configuration file, for example, the following data is stored:

  • Order lead time of the SC: two days
  • Delivery days of the SC: Tuesday and Thursday
  • Days when orders can be placed: Monday - Friday

Assume that today is Wednesday 07.08.2013. The next possible delivery date then according to the data stored in the configuration file is Thursday, 08/08/2013. Since, however, the stipulation for two days order lead time before this date does not apply, the date cannot be used for further planning. The subsequent possible delivery date is Tuesday, 13/08/2013. The condition for placing the order two days prior to delivery is satisfied for this date. Thus, Tuesday, 13/08/2013 is the next possible delivery date for which an order can be placed.
Since, as mentioned above, the latest possible order date is sought, the order date is now calculated from the determined delivery date, which is exactly two days (lead time of SC) before it. For the example given, this means that Friday 09/08/2013 is the latest possible date for ordering.
In order to decide whether an order should be placed, the latest possible order date must be compared to the current day to see if they match. In the present example the 07/08/2013 was compared with the 09/08/2013. Since these days do not match, the 07.08.2013 is not an order date.
This test is performed daily by the system. In this example, the next order day it would identify would be Friday, 09.08.2013.

Calculate Order Quantity

After the demand quantities and the order day have been determined, the order quantities for the order days remain to be determined.
To calculate the shipping amount, first the period must be identified for which the money supply should be requested. To work this out, data of the subsequent delivery are included as demand must be covered up to this date.
The example from the previous section is used below to illustrate the process of calculating the order quantity.
The day on which the order should be placed was Friday, 09/08/2013. The delivery date after next would thus be Thursday, 15/08/2013. The order amount must cover the period from the date of order until the following delivery date.


There are now two ways to handle this period. The determined period is counted either from the order date or from the delivery date. This is set in a configuration file.
From the example above there are thus two possible date ranges to determine the demand. The demand could be ascertained from 09/08/2013 to 14/08/2013 or from 10.08.2013. until 15.08.2013.
Now the previously defined quantities need to be used for further calculations.
To calculate the order quantity, add the demand quantities of the individual days of the determined period of time. The current stock in the ATM is subtracted from the total and the specified reserve in the configuration file is added.
This results in the following formula:

The concept presented here has been implemented in our CETIS CH product and contributes to the optimization of cash management. The demand for cash, which is related to the use of the system for cash management, can thus be easily calculated. This ensures that a minimum amount of capital must be used in order to ensure there is always a sufficient amount of cash.

Contact: Christina Wiesing; Turn on Javascript!