RPA, also known as robot process automation, also known as software robot, is the most popular software in recent years, the most popular software by enterprises.
People say it’s a subdivision of automation technology, a software robot, and others say it’s rubbing off on the heat of AI. However you define it, investors and markets love it.
The advantage of the RPA
Robot process automation (RPA) is developing with different features and advantages from traditional automation.
One: do not change the original IT system
RPA operates at a higher software level, which means that it does not invade existing software systems, but operates on them at the presentation level. Traditional automation relies on a technical infrastructure that needs to change to match the automated process. For example, Ping ‘an CLOUD RPA Anxiaobi, all applications and data are in the cloud, without deploying local computers and changing the original IT system, you can carry out software upgrades.
Two: small white also can get started quickly
Part of the RPA has been implemented so that users can use it without any programming skills, just by focusing on the functionality provided by automation. Traditional automation users require programming skills. Programming language requirements will vary depending on the automation tool used.
Three: Rapid deployment
Because RPA software is process-driven, it can be implemented from definition to launch in just a few weeks. In traditional automation, feasibility studies and trial design make this process take longer. For example, the RPA system of the head office of China Construction Bank went online on May 25, 2019. From bidding in November 2018 to implementation and launch, it only took six months.
RPA disadvantage
A: weak
RPA requires reconfiguring robots even with minor changes in the application. IT analyst Jason Bloomberg wrote in Forbes that the RPA’s main weakness is its poor robustness.
An automated robot that follows strict rules cannot quickly adapt to the many operations that have to be changed. Robots cannot adapt to changes in any aspect of the user interface, data, or applications.
Two: high failure rate
Deloitte found in a survey that of 400 companies using RPA, 30 to 50 percent of RPA projects failed at the outset, and 63 percent were not delivered on time.
Sanjay Srivastava, chief digital officer at Genpact, says: “In five years in the industry, there have been very few success stories in over 1,000 enterprise robot deployments.” “Robots need constant management and maintenance in production and life.”
The “Four Stages” of RPA
According to IDC’s research report, the global digital GDP will reach $46 trillion by 2022, accounting for 46 percent of the economy. As early as 2018, more than 50% of China’s Top 1000 enterprises have made digital transformation the core of their strategy.
RPA in the short term, but also with its “digital staff” attribute is inseparable. As a software robot, RPA can handle a large number of tasks on computers, enabling enterprises to transform digitally. The need for RPA far exceeds the need for traditional process automation.
In the process of exploration and practice, RPA has gone through four stages of development. (From the perspective of enterprise development)
- RPA phase 1.0: Automating individual businesses
At this stage, RPA covers almost all operations of desktop automation software. However, it failed to form a “closed loop” to achieve end-to-end automation of a certain business cooperated by several departments.
For example, the earliest RPA application in China: Button Wizard. Through the production of scripts, you can let the key wizard instead of hands, automatic execution of a series of mouse and keyboard actions. In the game can replace the player’s hands, automatic play strange, automatic blood tonic, automatic speaking; At work, it can replace all computer operations that employees can complete with both hands, such as automatic adjustment of document format, article typesetting, automatic sending and receiving of emails, etc. But keystroke sprites do not enable multi-department business processes.
- RPA 2.0 phase: Automating business collaboration across departments
At this stage, RPA can achieve end-to-end automation, making multi-department business collaboration automation a reality. At the same time, RPA robots can work 24/7 and replace human-machine interaction with business processes, freeing up more application possibilities. It is mainly deployed on VMS virtual machines and can arrange work content, centrally manage robots and analyze robot performance. The disadvantage is that the work of RPA still requires manual control and management.
Take a small practical case: a company doesn’t pay its employees bonuses on time. When problems arise, the human resources department feels aggrieved. They strictly follow the bonus payment process. The accounting department is also very aggrieved. When the data is sent to the HUMAN Resources Department, problems in the data are often not timely fed back to the accounting department. At this point, if you want to solve this problem, you must combine the two subprocesses into one large end-to-end process for analysis and solution. The second phase of RPA is to solve the problem of cross-departmental collaboration in automation, extending the scope of automation and solving more complex problems in the process.
- RPA 3.0 phase: RPA “on the Cloud”
In the third phase of RPA, RPA is typically deployed on cloud servers and SaaS and features automatic sizing, dynamic load balancing, situational awareness, advanced analytics, and workflow. The downside is that processing unstructured data is still more difficult and requires more powerful technology convergence.
The weak coupling of RPA enables low-cost and rapid deployment across software. With the trend of cloud-based enterprise services, cloud on RPA has become an inevitable result.
According to CompTIA, nearly half of all companies say 31 to 60 percent of their IT systems are cloud-based. Eighty-one percent of companies say cloud computing has significantly enhanced or moderately enhanced their automation efforts.
Cloud computing supports RPA with computing power, and RPA running on cloud is called SaaS RPA.
For example, Alibaba Cloud RPA has been implemented in the cloud, remote control and does not occupy the existing computer.
Cloud RPA is generally cheaper to deploy than development RPA and local deployment RPA. Because IT is stored in the cloud, there are no software clients and no venues, and enterprise IT personnel do not have to be involved.
From an IT perspective, cloud RPA software is always up to date, eliminating the need for upgrades on local machines and enabling businesses to seamlessly deploy the latest software.
- RPA 4.0 stage: RPA+AI
RPA is the best “cut” for an enterprise to enter AI. Using artificial intelligence, machine learning, natural language processing and other technologies, it can realize unstructured data processing, prediction specification analysis, automatic task acceptance processing and other functions.
Combined with AI vision technology (such as image recognition, face recognition, machine vision, biological intelligence recognition, etc.), RPA robot can recognize and screen pictures and videos, helping users to realize id card recognition, bank card recognition, automatic account opening of credit cards and other functions.
For example, in the credit field, there are many kinds of pre-loan audit materials, many formats and long length pain points. RPA such as philosophical data and cloud expansion technology can be extracted and reviewed based on OCR key information, supporting identification of id cards, loan ious, loan contracts and other photocopies, and enabling RPA with AI technology.
At present, most RPA software products are between 2.0 and 3.0. Some industry giants have begun to explore RPA 4.0 and have initially applied AI to enhance the cognitive ability of RPA products.
Gartner: The future of RPA is super automation
Peter Walker, Regional CTO for EMEA at Blue Prism, said, “Throughout 2020, RPA will be further ‘super-automated’ as a means of testing and deploying ARTIFICIAL intelligence, natural language processing, intelligent optical character recognition, communications analytics, process optimization, and machine learning deployments in the enterprise. And it’s becoming more and more popular.” Gartner’s top 10 Strategic Technology Trends for 2020 puts super automation, and RPA in particular, high on the list.
RPA is a good combination of software process automation technology and AI, two complementary concepts, so that high-quality white-collar labor force from repetitive, boring computer office work, so that they can devote time and energy to more creative work.
And the integration of RPA into a broader digital transformation strategy is a key element in achieving large-scale “super-automation”.
The original link: https://www.51rpa.net/rpanews/5377.html