Expectations for robotic process automation (RPA) continue to rise as an IT solution that will be immediately effective for increasing work style innovation and worsening labor shortage. RPA magically creates automation for existing operations without special programming knowledge. Many probably have already begun considering its introduction, but not a few may say that they have often heard about it, but that is it. We will examine why RPA is popular, which may be a question that may not be easy to answer at this point.
What is the Mechanism of RPA? Its Greatest Feature is Its High Versatility that Can Supporting Any Operation or Industry
There are three main reasons why RPA is highly popular.
First, it is an IT solution that can directly address the issues of work style innovation and labor shortages faced by office workers today. Company-wide bans on overtime have become common over the last couple of years. Time spent working in the office has physically decreased. Nevertheless, efforts to decrease the amount of work have been postponed in most companies, which makes increasing in work efficiency an urgent task in the office. Such companies would increase their manpower if they could afford it for the moment, but they are barely able to secure labor as it is, and even if they could, it is not easy to explain to new hires their company's unique internal operations and have them execute such operations smoothly. RPA can solve this problem promptly.
RPA is a solution in which software (called a "software robot") on a PC automatically executes daily tasks currently performed using PCs in the office. For example, let's say there was a task involving opening a Web page, capturing data, pasting it on Excel, and sending it to the person in charge. Previously, a person looked into a PC monitor while operating a mouse to execute this. If RPA is introduced, this task is performed by a software robot running on the PC instead.
Since the software robot proceeds with the operation autonomously without a break and works just like a worker, RPA is sometimes referred to as Digital Labor.
One of the characteristics of this mechanism is that it is not a solution that is limited to special industries or operations. Introduction of RPA can be highly effective for 1) tasks involving opening multiple windows on a PC and managing data between them, 2) tasks involving large volumes of data, and 3) tasks repeating the execution of a given sequence of operations. Productivity can increase significantly in any operation or industry.
Simple Introduction, Swift Verification of Effectiveness, and Elimination of the Risk of Human Error
The second characteristic is that the development of the software robot is easy. When developing RPA, the process involves 1) identifying the target operation, 2) creating the software robot, 3) implementation verification using a prototype software robot, and 4) feedback of the verification results. It is often the case that a software robot is completed after a week or two from the beginning of the process and put in operation.
The short development period has to do with the emergence of RTA tools for developing software robots easily. With RTA tools, software robots can be created easily even with no programming experience. For simple tasks, software robot prototypes can be created in two steps: 1) record the user's PC operations; and 2) add necessary operations to the flow chart of the recorded PC operations. A software robot is built quickly and used in the actual setting to identify points for improvement. Repeating this process can increase the operational efficiency while developing. This also allows users to do a partial introduction at first to determine the target operation and eventually greatly expand the target department and target operation.
The third characteristic is that it can free us from human error, examples of which include pasting retrieved data in the wrong place or sending it to the wrong destination. Particularly in operations repeating simple tasks such as entering large volumes of data, not only were people unaware of their making errors, but it was also not easy to check if the data was entered correctly. Another major attraction is that not only is work efficiency increased, but the risk of work errors is eliminated.
Entrust Low Profitability Tasks for Which Introducing IT was Difficult to Digital Labor
RPA, which is becoming a major boom, became introduced in earnest in companies during the last year or two. This year appears to be the first year of its widespread use. For example, Nikkei xTECH analyzes domestic RPA introduction based on the sales scale and order status of RPA tool vendors and market scale forecasts by research companies, and has estimated that the number of companies introducing RPA will exceed 5,000 by the end of 2018.
What operations/tasks are the early adopters of RPA using it for? Here are some typical uses as described in RPA Souran ("Overview of RPA") issued by ICT Innovation Research Institute, Nikkei BP Intelligence Group in March 2018.
- Entering the account opening data, etc. sent by e-mail from approximately 400 branches into a business system (Mizuho Financial Group, Inc./Mizuho Bank, Ltd.)
- Visit and login to over one hundred retailers' Websites to select and download required POS data (Sapporo Breweries Ltd.)
- Read payment statements with optical character recognition (OCR) and match it with information on money received (Teijin Frontier Co., Ltd.)
- Login to clients' systems and download the ordering data. Enter the shipment information, export the data needed from the system and convert it to an Excel file, and send it to the client as an e-mail attachment (Nisshin Seifun Group Inc.)
- Check regularly for unapproved forms on a backbone system and send an alert e-mail if one is found (Pado Corporation)
- Login to a customer database, update the information according to the rule, and log out (Telefónica O2, UK)
- Import data from dashboard cameras and create daily reports in Excel to automate the daily reporting task of approximately 700 company car drivers (Sumitomo Forestry Information Systems Co., Ltd.)
- When the staff reads the bar code for the policy number on the invoice, the software robot recognizes the policy number, and based on that, pulls up the relevant data from other internal system and automatically enters it into a business system (Nippon Life Insurance Company)
If we look at these operations adopting RPA, we can see that they are operations from areas in which making IT investments was difficult due to such investments not being directly connected to profit, or low profitability. It is possibly the case that, operations with low profitability are being entrusted to digital labor, allowing employees to spend more time on creative work with higher profitability
What is Important in Implementation is Figuring Out the Applicable Operations and Maintenance After Introduction
The most important key to successful RPA introduction is by far figuring out which operations it can be applied to. Using RPA tools, a software robot of some type can be made for any operation. However, not all software robots can achieve significant increases in productivity. First, the operations must be identified and the PC tasks analyzed for each operation, before carefully examining the degree of improvement in productivity that can be achieved when the PC tasks are executed by the software robot. If there are operations that may lead to an increase in productivity, the PC tasks should be analyzed and a software robot prototyped. Such a robot should be partially introduced to measure its effectiveness and identify problems. The software robot's level of completeness should be increased before proceeding with full introduction.
One thing that tends to be forgotten is that created software robots require maintenance. The management system for the software robot should be organized based on the premise that the destination e-mail address may change or the URL to be accessed may be changed frequently. Specifics on how to conduct maintenance should be decided and periodic reviews should be included in the operation.
Example of RPA Utilization: The Key is Not Relying Entirely on Computers for Its Operation While Clarifying the Content of the Work
What are the keys to utilization when actually operating RPA? Let's look at an RPA utilization case study at Fujitsu Communication Services Limited (CSL).
CSL uses RPA for back office operations, applying it for inspection and cancellation operations. With the introduction of RPA, robot software can find errors in work when they occur, freeing workers from daily pressure. During a peak period, the amount of work increases by five or six times and they had difficulty securing personnel. With RPA, they can now respond without increasing staff.
The first key in RPA utilization at CSL is creating a hybrid of human and RPA. Rather than relying on computers for everything, users ensure that there is human intervention. This is to clarify who is responsible when problems occur. To ensure this, CSL assigns a person in charge for every PC RPA is running on.
The second key is to clarify the content of the work when organizing the operation. Good results cannot be expected unless this is done thoroughly. CSL was able to shorten the time taken for and streamline new employee training.
After the appropriate introduction and operation of RPA, feedback was received saying things such as it was now possible to label the difficulty of operations, it would be now possible to employ many workers with disabilities, taking time off had become easier, employee training had become easier, and employees could leave work earlier and spend time on themselves. CSL was also able to confirm with questionnaires that they now have a very friendly work environment.
The way in which digital labor is utilized will differ depending on the kind of work style innovation that companies and organizations are aiming for. In any case, the key to the success of work style innovation depends on whether or not we can include digital labor as a reliable and outstanding fellow worker.
Chief Research Officer
NIKKEI BP Intelligence Group
Graduated from the faculty of Engineering, Tohoku University in 1985 and joined Nikkei Business Publications, Inc. in the same year. As reporter and assistant editor for Nikkei Data Pro, Nikkei Communication, and Nikkei Network, covered and wrote about advanced technologies related to communication/information processing as well as trends in standardization/commercialization. Chief Editor of Nikkei Byte in 2002, Chief Editor of Nikkei Network in 2005, and Chief Editor of Nikkei Communication in 2007. Publisher of ITpro, Nikkei Systems, Tech-On!, Nikkei Electronics, Nikkei Monozukuri, Nikkei Automotive, etc. before becoming Director of the Overseas Business Division in January 2014. Has been in his current post since September 2015. Has been writing the series Jido unten ga tsukuru mirai (A Future with Self-Driving Cars) since August 2016 on Nikkei Online Edition. Issued Sekai jido unten kaihatsu project souran (An Overview of Global Self-Driving Development Projects) in December 2016 and Sekai jido unten/connected car kaihatsu souran (An Overview of Global Self-Driving/Connected Car Development) in December 2017. Serving in the CEATEC Award Review Panel since 2011.