THE IMPLEMENTATION OF ROBOTIC PROCESS AUTOMATION (RPA) AND ANALYTIC HIERARCHY PROCESS (AHP) FOR PROJECT PRIORITIZATION IN BANK SASA

• Main Author: Haikel, Syem
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/63568
• Institution: Institut Teknologi Bandung

Robotic Process Automation (RPA) is a technology that configures a “robot” to emulate and integrate the actions of a human interacting within digital systems to execute a business process. It was an innovation in Bank SASA for digital transformation. The primary purposes are efficiency and mitigating risks to the bank because most banking transactions are repetitive tasks and have high transaction volumes. It takes much time consumed and could increase the probability of human error. Bank SASA’s managements wish that RPA can replace their employees to do those tasks and let them do more strategic and analytical jobs. To implement RPA technology, Bank SASA established a department at the beginning of 2020, namely the RPA Office, to focus on implementing RPA. This research wants to solve two main problems regarding the current business situation in SASA. First, the implementation of RPA is not optimal enough or did not reach its full potential yet. Since the first implementation at the end of 2019, much improvement is still needed. This research uses a fishbone diagram to analyze and find the possible causes of that problem. The author uses six categories: people, method, measurement, environment, preparation, and machine. The results are 20 causes among those categories. This research wants to solve one of the causes, which is not standardized prioritization. The author uses one of the decision-making tools, which is the Analytic Hierarchy Process (AHP). AHP was developed by Thomas L. Saaty. As one of the decision-makers, the author needs to determine the goal, criteria, sub-criteria, and collect the available alternatives, along with the other three decision-makers. The goal is to prioritize RPA projects. There are five criteria and six sub-criteria. By conducting pairwise comparisons on criteria and sub-criteria, the highest local weight for criterion is efficiency, 0.473. Meanwhile, the highest global weight is cost reduction, 0.228, part of the efficiency criterion. After RPA Office conducted socialization to some departments, there were 22 selected projects as alternatives. Instead of pairwise comparison, the author uses rating for selecting the best alternative. There are two to five categories on each criterion or sub-criterion. The top priority goes to New E-Tax alternative with a 0.78 weight value. The second main problem is related to the delay in disbursing the funds from the insurance company. It is one of the tasks conducted by a unit in SASA’s Operations Directorate. The author uses the problem tree tool to analyze and find the problem’s cause-effect. There are 12 causes, such as high transaction volume, many human errors, and consuming much time. Based on those causes, the solution is automating the process by using RPA. The project was called Insurance Claim. There are some required steps in implementing RPA. It starts with the general assessment regarding the existing process, then calculates the project’s cost and benefit. The result shows that SASA will gain loss in the first two years, but it changes to benefit for the next three years. The author uses a five-year projection, resulting in 28.6 million cost savings within five years. The costs will reduce to zero for the following years because all the costs are one-time payments at the start of the project. Then RPA Office conducted a further assessment and simulation regarding the existing process. That was needed to analyze and design the new process if robots replace humans. The robot replaces HO Officer as Process Owner in this project. The last steps before the implementation are robot development and testing. There are four positive test cases and three negative test cases that have been successfully tested. The end of this research is at the testing phase because of some reasons which affect the delayed implementation.