Keynote - Robotic Process Automation vs Micro Services13th of May 2019 |
Sometimes Ordina likes to provoke its experts. Every new project launches a discussion no holds barred about approach and technology. For example, what is the best technology for a company that wants to innovate at an accelerated rate? Robots to the rescue or perhaps services? During the first expert night, we put this topic to our experts: Micro Services adept Peter De Kinder and Tom De Maeyer, who already has 160 software bots on his resume. Both successfully braved the questions fire of colleagues, customer and enthusiasts.
Digitalization puts companies on a rollercoaster ride. Those who don’t want to stay ahead, need to adapt to an everchanging market at a breakneck pace. This can’t be realized by leaning on the monolithic software platforms of the previous decades. Companies are actively looking for alternatives to this approach. An increasing number of companies are automating repetitive tasks using Robotic Process Automation (RPA). Giants like Amazon and Netflix (and many others in their wake) are embracing micro services wholeheartedly in order to work agile.
Tom starts off: “Even the most successful IT implementations often leave a bitter aftertaste, because of their development time or because of considerable aftercare costs. Even though time-to-market is of such importance nowadays. RPA offers an answer to this problem: Software bots are experts at executing repetitive, standardized tasks. They are ideally suited to automate activities involved numerous applications. Their speed and accuracy is unparalleled and they can be swiftly implemented to get a fast return your investment.”
“Business Value is only part of the story” replies Peter. “Can we also adapt robots quickly when needs or organizational structure changes? Micro Services replace the classical client-server architecture backed up by a database and mirror their architecture on the existing business domains. You construct small pieces that meet a certain need. The services communicate with each other to achieve goals that used to be enveloped by a single application. When changes are needed, there is no longer a need to adapt the entire platform. You just swap out the micro services that are involved, and the rest is unaffected.”
“Also, Micro Services can be built in any technology.” Peter continues. “Adaptability, flexibility and scalability can be achieved within impact on the other components in your IT landscape. This allows you to react rapidly when needed.”
Tom disagrees: “Micro Services require a fundamental shift in thinking: You need to split up in small functionalities. This shift requires some discipline and time, no? Choosing RPA allows for a more pragmatic approach. After only four to five weeks your first bot can be operational. Although I must admit that adaptability has long been the Achilles heel of RPA, but this has drastically improved since its conception. Adapting a bot will take anywhere from a few hours to a few days, depending on the number of involved applications. On top of that, a strongly distributed environment allows for easy scalability: You activate a new bot, provision its account, and it’s off to the races. This is one of the reasons RPA is often used to replace a temporary human workforce.”
“Concretely, Software robots seem to me like they are meant for pursuing Operational Excellence”, Peter replies. “In theory there are three different strategies a company might pursue: Operational Excellence, Customer Intimacy and Product Leadership. Those who really want to innovate, are better of with Micro Services. Since adaptability is one of their greatest trumps, micro services are best suited for companies that want to distinguish themselves by launching new products quickly or want to be helped in achieving exceptional customer service.”
Tom reacts: “I am convinced that robots can have a positive impact in these areas as well. Take customer service for example: Every person will answer a customer question differently. With bots you can have a uniform – the most suited – tone of voice, scenario, etc.…”
Tom: “People are indeed scared that Robots will take their jobs. But isn’t this the case for any IT project? Aren’t cost savings are the main goal of any IT project? In reality, bots almost never lead to firing people. People are just reallocated to more strategic areas: new communication channels, getting to know the customer better, thinking about new business models and products.”
“With Micro Services you encounter this type of resistance much less” says Peter. “As previously stated, there is however a need for a cultural change. And Micro Services do require developers, since we are talking about custom development. But Micro Services are a long-term investment: Once they are up and running, you quickly earn back the investment.”
Peter: “Micro Services are technology agnostic. You can easily build one service in Java and another in .NET. Through the API they can communicate effortlessly. They are often even deployed in the cloud rendering them infrastructure independent as well. In that respect there is ample flexibility.”
“There are lots of software products to implement software bots. They are gaining in maturity and more and more consolidation is taking place.”, says Tom. “We compare the major players and they are very similar. It’s hard to choose unwisely this way when it comes to quality. Their philosophies are different sometimes: Either focusing on complete automation of tasks or focused more on interaction with people.”
Peter: “Ordina did a very successful project with Micro Services at Materials Handling giant TVH. Their existing applications could no longer sustain the rapid expansion of the company in an efficient manner. Our DevOps team replaced the legacy code with Micro Services in the cloud, each taking care of part of the overall functionality. Their organization today has gained much in terms of adaptability.”
Tom: “In the past I set up a multitude of robots for a large bank, for example to facilitate information retrieval from several sources. In total the work of 17 full-time equivalents were automated for an investment of around 50K €.
“As a company, why not invest in both disciplines?” This question was on the lips of several attending people during the discussion. Peter and Tom agree: “I believe they are two strong solutions for different needs. Is your focus the automation of process, go with RPA. Does your architecture need to be flexible because of fast paced market changes? Micro Services are ideal. Evidently there are other factors such as investment, implementation time, resources and such that also play a part in the considerations. But it is perfectly possible to integrate RPA with Micro Services.”
In short: The battle of this expert night did not have clear cut winners or losers. Maybe they are a match made in heaven?
Review | Keynote | SOA | BPM |