The conditions for implementing robotic process automation

Conditions for the implementation of an RPA The candidate processes for the implementation of a basic RPA must present the following characteristics, in two axes. On the one hand, the terms of reducing the human burden and the return on investment must be taken into account. In this case, they must have a strong but low-value-added manual component, be routine and repetitive, and in large volumes. On the other, the technical feasibility must be observed taking into account that the processes are digitized, that they are standardized and with few variations, based on rules and that they are stable and long-lived. Software robots replace humans by reproducing their actions, to relieve them of repetitive, boring tasks, without intellectual interest and with low added value. The stronger the replaceable manual component, the better the return on investment, by allowing repositioning of workers on higher value-added tasks or by eliminating an outsourced service.

• Routine and repetition

Candidate processes must be activated routinely and / or repetitively, in the latter case, the more frequent the better. The obvious reason is the return on investment. The more a robotic solution replaces humans, the faster the return on investment.

• Large volumes

In relation to the frequency and volume of work mentioned above, the processing and manipulation of large volumes of data, files, number of tasks, etc. completes the robot portrait of the candidate processes.

• It is likely to produce errors

Highly repetitive, high-volume, and unattractive processes are more likely to cause errors. In fact, this type of activity can cause boredom, inattention or present special difficulties, particularly related to the speed required to complete tasks when processing large volumes. Errors produced can negatively impact customers, company image, process performance, require corrective actions, possibly lead to a loss in billing, margin or even the application of penalties.

• Digitized processes

The candidate processes must already be digitized, be compatible with a software robot. The processes that are still linked to paper and pencil are technically incompatible. Still, it is not uncommon for a factory manager to handle paper forms or manage filing cabinets to store documentation.

• Standardized, with few variations

For the implementation of a simple RPA, the planned processes must be standardized and not present variations. Standardization involves the description of the stages of the process that are observed by the performers. A process whose execution is linked to personalized knowledge or that varies between performers must first be standardized. This standardization benefits from being based on best practices, that is, from providing the best overall efficiency in execution. The variants in the processes must be taken into account by automation through the creation of different robots or more complex and intelligent robots, capable of handling the different variants. If variations appear over time, it may be necessary to edit the scripts or create variations. This negatively influences ROI and potentially reintroduces digital complexity. It would be paradoxical to set up an RPA and have to manage the variants with human actors, which could raise questions about the relevance of the initial use of RPA.

• Rule-based

In addition to standardization, candidate processes must be rule-based and transferable to software robot programming. These can be confronted with options during the execution of the script. The more this type of choice is based on rules that are unambiguous and open to interpretation, the simpler and more reliable the RPA programming will be.

• Stability and longevity

The candidate processes must be stable and have sufficient longevity to justify their automation. Stability means little or no update needed over a long period of time, for example a year after automation starts. Longevity or sustainability is the predictable life of the process. If it is likely to disappear before a long period of time, which may be 2 years, this should be taken into account when evaluating the relevance of passing the process under RPA and, if necessary, to calculate the return on investment. A process may disappear due to technical development, reengineering, regulatory change, and contractual or regulatory requirements.]]>