The requirement was to evaluate the potential risks associated with the design of a cash handling mechanism to provide confidence that the specified performance could be achieved.

Handling bank notes and other "documents of value" in automatic mechanisms gives rise to a particular set of potential problems. We needed to provide a way of systematically documenting and analysing the main risks in the mechanism design and to show how these risks could be minimised or, at least, managed.

We chose to use the Failure Modes and Effects Analysis method as a well-proven approach to risk identification, prioritisation and reduction.


The mechanism consisted fundamentally of a series of banknote transport and storage systems and was used to validate and store customer deposits. The lifecycle of documents through the machine was plotted so that none of the major risk contributors would be missed. By dividing the mechanism into sections bounded by risk "hot spots", it was possible to define a set of segments that would undergo evaluation. Each of these segments was addressed in turn by listing all of its possible failure modes and then, for each failure mode, recording the "Severity Rating", the "Detection Rating" and the "Occurrence Rating". The product of these three ratings gave an overall "Risk Priority Rating" for that failure mode.

Severity Rating (SR)

The severity ratings of failure modes were defined in terms of the perceived impact that the risk would have if it were realised. These ratings range from 10 (most severe consequences) to 1 (least severe).

Priority Area affected by fault Severity
1 Safety 10
2 Security against banknote and data theft 9
3 Regulatory compliance 9
4 Documents are transferred between owners 8
5 Accountancy error against customer-owned notes 8
6 Media damage (customer’s notes) 6
7 Machine out of service 6
8 Reject rate 4
9 Media damage (bank’s notes) 3
10 Mis-stored notes 3
11 Throughput performance 1
Detection Rating (DR)

The probability of detecting that a certain failure mode has occurred was then defined. If there was no way of sensing that a failure mode had occurred this was rated at 10 and conversely, if a failure mode was almost certain to be detected this was rated at 1.

1 No chance of detection
2 Very remote chance of detection
3 Remote chance of detection
4 Very low probability of detection
5 Low probability of detection
6 Moderate probability of detection
7 Moderately high chance of detection
8 High probability of detection
9 Very high chance of detection
10 Almost certain chance of detection
Occurrence Rating (OR)

The potential occurrence rating was defined in terms of the time that the system may run between fault occurrences and from that we could deduce the number of documents between occurrences of the particular fault. The assumption made here was that, on average, around 15,000 documents were processed per week.

Definition of frequency in running time Probability in number
of documents
10 Almost inevitable – there is a 1 in 2 chance of the fault occurring in one day > 1 in 1000
9 Very high – once per day 1 in 2000
8 Repeated failure – once per week 1 in 15,000
7 High – once every 2 weeks 1 in 30,000
6 Moderately high – once per month 1 in 60,000
5 Moderate – once every 3 months 1 in 180,000
4 Relatively low – once every 6 months 1 in 350,000
3 Low – once per year 1 in 700,000
2 Remote – once every 2 years 1 in 1.5 million
1 Nearly impossible – less than once in 8 years < 1 in 5 million
Section Failure Modes

The following table gives an example of some of the failure modes identified in a section that includes a mechanism for separating notes.

SR = Severity Rating
DR = Detection Rating
OR = Occurrence Rating
RPR = Risk Priority Rating
Where RPR = SR*DR*OR.

Item Failure Mode SR DR OR RPR
1 Failure to feed notes 1 1 9 9
2 Stream feed of notes 4 1 6 24
3 Skewed feed of notes 6 1 6 36
4 Note damage (fold, tear) induced 6 1 8 48
This process was repeated for all failure modes and for all sections of the machine giving a complete picture of the risks associated with the mechanism design. Those failure modes with the highest Risk Priority Ratings were given the most immediate attention and design changes were identified to either, reduce the probability of occurrence or, to increase the probability of detection so that it would be possible to take remedial action whilst in operation.

Using a technique such as Failure Modes and Effects Analysis allowed us to systematically identify the risks in a mechanism design, quantify those risks with relative ratings and recommend design changes to increase the probability of product success. This process was used as a very effective tool for verifying if the proposed layout and design of the mechanism adequately coped with the failure modes that would predictably occur during machine operation.

Call Me 
Case Study
Web Enabled
Document Handling
Document Handling
Mechanism Control
Mechanism Control