Robotic Process Automation (RPA) has become indispensable to companies: productivity, customer satisfaction, employee development, quality and compliance, lower operating costs… These are the essential challenges that Robotic Process Automation, whose tools have only recently matured, can meet.
RPA is a rapidly developing technology because it is closely linked to the acceleration of the digital transformation. As users seek a better digital experience, their behaviour is changing. In response, companies are also forced to adapt their relational and operational models. For what results? Aligning the level of service with user expectations, creating value, personalising relationships, reducing costs, digitising processes and refocusing employee activity on added-value tasks.
Automation can thus streamline various types of processes:
- Manual, repetitive and time-consuming tasks in the front and back office;
- Multi-application, multi-data-source tasks requiring interaction with all types of tools;
- Structured tasks, with rules and predictable decision logic, etc.
These processes are automated via the development of robots capable of carrying out numerous operations, such as connecting to applications, processing and sending e-mails, reading and writing in databases, performing calculations, etc.
ALTEN has designed this approach to serve its customers around four key RPA implementation stages.
1. Initiating the process: A crucial phase
This phase is absolutely essential because it allows us to define the customer’s expectations and to identify the automatable processes that best respond to them. It is structured around three major stages:
- Mobilisation and awareness-raising to get the teams on board with the RPA approach, based on collaboration;
- Identification of processes eligible for automation, via an eligibility matrix that allows their evaluation according to certain parameters (Benefits vs. Complexity);
- Selection of processes according to the criteria defined by the client and the pre-evaluation.
To begin this phase, it is vitally important to present the added value of the RPA approach to the different teams: scope, eligibility criteria, progress of the main stages of the project, REX and RPA demonstration targeted at the audiences present, etc.
Working closely with each team, particular attention must then be paid to the customer’s expectations, which the RPA approach must be able to meet: quality, productivity gains, compliance, improved customer satisfaction, etc. These expectations will make it possible to weigh the expected benefits against the associated complexity/development workload, in order to define the processes that will be good candidates for automation.
The construction of an eligibility matrix (in particular the one developed by ALTEN) makes it possible to equip this approach and to simply and quickly highlight the best candidates for automation, based on a trio of actions:
- Evaluation of each candidate process against eligibility criteria (operating rules, type of process, data structure, frequency of change, number of exceptions, frequency of errors, type of applications, etc.);
- Automatically assigning one score on expected benefits and another on complexity of development;
- Restitution, based on the results obtained, in the form of a Quadrant distinguishing four priority levels of development, which make it possible to highlight the Quick Wins (high benefit and low complexity).
A prioritised backlog and implementation schedule for the initial automations are the main deliverables of the initiation phase. It is the essential basis on which the RPA project can be structured and successfully deployed. Introducing a pilot or PoC into the process provides feedback for optimisation and operational adjustments. This initiative also gives the client a quick overview of the added value of the approach.
2. Defining the governing principles for the deployment of robotisation
The second phase of implementing an RPA project starts with defining the target organisation that will offer an optimal deployment of the approach.
A preliminary workshop presentation of the different potential organisational models should enable the stakeholders to plan for the one that best suits them, while ensuring the top level of service.
This definition leads to the description of the Target Operating Model (TOM) of the RPA project, which must fulfil all the application conditions for the implementation of the approach. The range covered must allow the documentation of the main areas impacted by its implementation: roles and responsibilities on the life cycle of the automation request, change and performance management, governance, property areas.
Finally, stakeholders’ often legitimate questions about changes in the nature of their jobs or the evolution of certain business processes make it necessary to initiate and carry out a long-term change management plan whose objectives may especially:
- Emphasize the added value of the RPA approach;
- Provide insight into how processes eligible for automation will impact their nominal mode of operation;
- Build a training plan and prepare each stakeholder for their new responsibilities in the RPA deployment;
- Raise awareness and mobilise staff around a real business project so that everyone becomes an important player contributing to the project’s success.
3. Automating the manual process
Once the target organisation has been defined, robot design and development activities can be initiated, with the aim of taking a perfectly described manual process and matching it with an automated process. What are the main steps to achieve this?
- Information gathering (process metrics, nature of applications, key features, infrastructure requirements, etc.);
- High-level modelling;
- Definition of input and output data of the process;
- More detailed modelling of the process, identifying and describing the different stages;
- Identification and handling of business and system exceptions;
- Production and validation of the Process Design Document (PDD), which gathers all the information and whose completeness determines the quality of the robot’s development.
Once the PDD has been validated, the technical specification stage can begin, with the main deliverable being the Solution Design Document (SDD), which will serve as a reference for the development of the robots. It can be done in “agile” mode to get intermediate validations and guarantee the involvement of all stakeholders.
As with any solution development, after unit testing and technical integration, the robot is delivered for acceptance. This is an essential step, as these robots run in production environments and any incident can lead to major malfunctions. It is therefore imperative to check that the developments correspond to the needs expressed and that they meet the requirements of robustness.
Key success factors for this stage include:
- A good test strategy which describes the scope covered by each stakeholder (IT, business, etc.) and defines the output conditions of each level of testing;
- Test environments that can accommodate and bring together all the application layers that the robots pass through;
- Definition of the data sets needed as input to the process (list of transactions, user accounts, etc.) in order to achieve an optimal valuation that allows the different scenarios (nominal, alternative, etc.) to be run through;
- An acceptance schedule, describing the design and execution periods for each level and specifying the duration and intensity of the work of the various players (POs, business lines).
4. Sustaining and mainstreaming the approach
The final stage of the approach designed by ALTEN first involves ensuring that the robots are maintained in their optimal conditions of use, their evolution in the event of a change, and the integration of optimisation channels from the ongoing improvement plan.
The range of improvements adopted should be as broad as possible to ensure optimal performance in the following areas: robustness, processing time, execution ratio, organisation and orchestration of robots, etc.
Particular attention should be paid to the change. Indeed, any change, even a minor one, in one of the applications that make up a process path can potentially have an impact on the operation of the robot. It is therefore good practice to build a traceability matrix that allows a link to be established between the changes contained in the input change request and the impacts on the various components. This system will make it possible to effectively ensure the maintenance of robots in operational conditions.
Establishing a support structure and activities is also essential for sustaining and mainstreaming the RPA approach. The effectiveness of these activities is conditioned in particular by the establishment of a documentary repository constituting the knowledge base (user manuals, specification documents, production start-up procedures) and the definition of roles within the various levels of support implemented in the organisation. Finally, if certain tasks are to be taken over internally, a skills transfer plan must be implemented. Progressive support, supplemented by a training plan, will ensure that the teams’ skills are developed in various areas, until they become independent.