In their daily lives, purchasing and accounting teams are required to carry out many time-consuming and repetitive tasks, usually tedious. To relieve them and free them up time, RPA technology allows some of these tasks to be entrusted to software robots, who perform them continuously and reliably. This innovation was the subject of the webinar organized in mid-February by Acxias and its partner UiPath, followed by some forty purchasing, procurement and finance decision-makers, illustrated by the testimony of the Safran group.
You are responsible for purchasing or finance, and you want to free your teams from repetitive and time-consuming operations? The solution is there. Its name: Rpa, for Robotic process automation, in other words, robotic process automation. "The introduction of software robots, which a vast majority of companies consider to be one of the main levers of their digital transformation, addresses several challenges: increasing revenue and lowering costs, improving customer satisfaction, and reducing the risk of non-compliance," said Ludovic Duverger, Technical Partner Manager at UiPath, as an introduction to the webinar. According to him, delegating all tedious tasks to robots also enhances employee engagement, which can focus on "what really matters" and enjoy a better quality of life at work.
But how does it work? While a mini-express survey revealed a good level of maturity on the subject, the head of UiPath went into detail about the Mechanisms technology. The RPA associated with Artificial Intelligence (AI) allows a collaboration between the digital and the human to delegate some of the repulsive tasks (seizures, copy and paste, information searches, etc.) to robots. It can be an "assisted robot", a kind of personal digital companion available on each station, which the user can request to help him on a daily basis, or "un assisted robots", installed in the back office and working without human intervention except to manage the exceptions. In practical terms, "these robots mimic human actions, with any type of application, to execute an operation or manipulate structured data, continuously and without making mistakes," said Ludovic Duverger, adding that this type of project is quick to deploy: from a few days to a few months at most in the most complex environments.
For companies convinced of the interest of RPA, the challenge is to identify potentially affected processes and then develop robots by setting priorities, according to a rigorous methodology, such as that presented by Acxias. "A good candidate is a long, cumbersome and repetitive process, which represents a certain volume to guarantee a good return on investment," according to Guillaume Vassaux, head of integration and innovations within the agency in digitalization. "The process must also be fully controlled and well documented, to know how to manage potential exceptions." To present the methodology, he relied on several cases of use responding to problems encountered during various software implementation projects, in particular SAP Ariba. First, the registration of suppliers, the creation of tenders and the updating of contracts, by recovering or replication data from different sources (ERP, mail, etc.). Then theuser administration, based on requests for change. Finally, control purchase requests and, by adding AI features, create them automatically from quotes.
To illustrate the control of purchase requests, Safran presented the robotic automation project that allowed it to offload its buyers from tens of thousands of manual interventions per year. Before the implementation of robots, a preparatory phase had identified eligible processes and validated automation on a specific perimeter: requests of small amount out of catalogue containing a quote in attachment, with the aim of being able to transform them into orders without any human intervention other than their validation. Technically, the process revolves around two robots, developed by Acxias. The first analyzes the quotes of the purchase request, to extract the information and return it in a structured format. The second robot reads this structured information and compares it to the demand data in SAP Ariba, before validating it. Today, "the work of robots, which operate continuously, allows us to absorb peak loads," said Idriss Mrani, head of the Operational Support Department (DSO) in the group's purchasing department. Buyers no longer have the pressure they need, and end customers see no difference because robotics are completely transparent to them, other than an acceleration of treatments.
More information: report (long version), replay, presentation slides