Robotics process automation - saving costs, one mundane task at a time
By Hong Kong Economic Times

When Chief Information Officers on the cutting edge of technological adoption talk about the benefits of business process automation, the ensuing reaction may not always be that of awe and wonder. “Will automation replace my job?” employees in lower-level operation roles might wonder. Even other C-suite executives too may doubt: “Will our investment into automation fail deliver on its promise to cut costs?”

While there is a chance that both worries and doubts may turn out to be true, many success cases across multiple industries indicate that the benefits of automation, in this case robotic process automation (RPA), have become a necessary tool for staying competitive. McKinsey & Company’s study into RPA adoption across 16 case studies from varied industries showed that companies were able to achieve an average of 30% - 200% return-on-investment (ROI). Of course, each company’s ROI is defined by its unique needs and dynamics - for some, ROI can be determined by man-hours saved. For others, success may be having shortened lead times for overall project deliverables.

No matter what companies may try to achieve with RPA, it is clear that such automation is far from being simply a passing fancy. According to global research and advisory firm Gartner, the RPA market will reach $1 billion by 2020, with 40% of large enterprises having adopted RPA by next year (up from 10% since 2018).

More locally in Hong Kong, a joint survey by ACCA Hong Kong and KPMG China shows that the three main reasons for investing in automation technologies in Hong Kong include: improving efficiency and prioritising value-adding tasks (29%), minimising human error in manual processes (25%) and cost savings (20%). That being said, RPA adoption in Hong Kong is still in its beginning stages, with 80% of survey respondents saying they have never even heard of automation technologies and have no plans to adopt them. This survey interviewed 388 respondents from government, public administration, finance, manufacturing, professional services, retail, insurance, real estate, IT and telecommunications.

More importantly however, is the question of whether RPA holds any value for companies in the field of engineering. A simple answer to this would be an unequivocal, confident yes. As we will see some engineering-related case studies, maintaining compliance and accountability with vendors in all areas throughout the construction process benefitted from RPA.

Breaking down RPA
RPA is a type of business process automation, but not the kind of “robot” you might find on a manufacturing assembly line, and not the walking, talking, doomsday-bringing kind you would see in the movies. RPA is better described as a “software robot” that users can programme to perform rules-based business processes using structured inputs. These tasks include things such as:

Data-entry tasks, such as importing information from one database to the other.
Webscraping, or finding information from online sources and compiling them into a format set by the user.
Cross-referencing data sets on command and highlighting inconsistencies between the two. This is very useful for monitoring documents and invoices for compliance issues.

This software functions at the user interface (UI) level, meaning it will perform tasks on a computer just as a human would. For example, if your daily routine consisted of literally copy and pasting numbers from one spreadsheet application into a report template in another application, you could programme a software robot, or “bot” to do this work for you - provided that all the fields and in which the variables go remain the same. Generating automatic responses to emails using data pulled from multiple applications is another time-saving way to streamline business/customer communication.

In an interview with McKinsey about its study on RPA implementation in different industries, Leslie Willcocks, a Professor of Technology, Work, and Globalisation at the London School of Economics, describes it this way: “RPA takes the robot out of the human. RPA is a type of software that mimics the activity of a human being in carrying out tasks within a process. It can do repetitive stuff more quickly, accurately and tirelessly than humans, freeing them to do other tasks requiring human strengths such as emotional intelligence, reasoning, judgment and interaction with the customer.”

Employees can consider applying RPA into any business process that is repetitive, rules-based and that doesn’t require judgment calls to be made by a human. Exceptions to the workflow can and will occur however, leading to potential errors in the bot’s output. While additional rules can be set to “teach” the bot what to do when it encounters exceptions, these rules must be added every time a new exception occurs. The more you teach the bot to jump these hurdles, the more your bot can automatically deal with similar scenarios in the future.

Here’s an example of one section of a bot’s workflow:

This is the workflow of a bot programmed to automatically process invoices from suppliers and vendors. Note that exception-handling is programmed into the bot’s workflow, which means that humans are still an important factor in keeping the system from derailing. The “forks” of the process in red colour account for if/then situations, such as in the case of an incorrect invoice.

Where is RPA currently being used?
Regina Viadro, Vice President at EPAM Systems, says that COOs at finance companies were among the first to adopt RPA with the goal of streamlining business processes without increasing costs or manpower. Deloitte LP enhanced its claims process by automating 13 processes using 85 bots, which handles 1.5 million requests per year. From this, Deloitte added the capacity equivalent of 200 full-time employees at the approximate cost of 30% of the cost of recruiting more manpower.

In McKinsey’s study, they worked with an insurer that cut down the time it took to handle 500 premium advice notes from two days to 30 minutes. While processing an advice by itself is far from a complicated task, it is a task that requires human effort. With mundane processes like these, bots speeded up the processing time produced accurate and consistent results. This saved employees time from backtracking to solve problems due to human error. “Now a lot of that sort of work can be automated” says Willcocks. Not every part of the process can be automated though, and some “requires human intervention, human reasoning, judgement. A RPA engineer would look at that type of process and say “which bit can we automate? The answer is, ‘Not everything’ - it can’t structure data.”

Here are some additional examples, aggregated by AI multiple, of companies using RPA and the results it brought them.

More locally, there is much room for local RPA cases to grow in Hong Kong. According to ACCA Hong Kong’s survey, only 4% of respondents noted that their finance departments were highly automated. Perhaps awareness is the main issue here, with only 13% of survey respondents saying they needed to invest more into automation solutions now to acquire first-mover advantage. However, 33% said they would consider investment in automation technology in two to three years’ time once the technology became more mature. The survey also revealed the three main reasons for investing in automation technologies in Hong Kong: improving efficiency and prioritising value-adding tasks (29%), minimising human error in manual processes (25%) and cost savings 20%).

RPA benefits for engineering firms
So far, you may have a good idea of how RPA can be beneficial to organisations across a wide spectrum of industries. Many of the benefits we have seen with RPA from these diverse companies also apply to those in engineering, of which include:

• Higher accuracy by minimising human error
• Faster, more streamlined processes
• Scalable bots - subscribe to more if you need them, unsubscribe if you don’t on a whim (in an RPA-as-a-service model)
• Reducing staffing costs and man-hours
• Increased staff productivity and freeing manpower to perform higher-value tasks

More specifically to engineering however, are benefits like monitoring cost and maintaining compliance for saving future headaches - things that decision-makers at engineering companies should consider RPA a viable business solution.

A report by Ernst and Young for the Government of the North West Territories in Canada explains how an RPA strategy can enhance the processes related to monitoring progress, vendors and compliance in large and complex capital projects. In this strategy, bots are used to calculate and track project cost data, which is then used to create reports for project management, supply chain management and the finance department to underline errors in compliance. We elaborate on some of these benefits below:

Particularly in the area of compliance with regulations and accountability with vendors, bots make the job of cross-referencing and highlighting inconsistencies in progress reports much faster.

A human agent performing tedious compliance work has a higher chance of making a mistake than a bot does. This can lead to increased processing times that come from having to backtrack in order to resolve issues. It also notifies employees of errors in a timely manner, which means less chance of unaddressed errors creating more problems as the project rolls on.

Cost efficiency monitoring
Being able to track project cost data and progress easily also gives management increased visibility into vendor performance and assurance for project success. Assuming the monetary value gained from the completion of a project is usually lower than the expected value as reported before the project began, RPA can help lessen this gap by giving management accurate, up-to-date cost information to put a stop to unnecessary costs. This, along with the many other benefits of an RPA enabled enterprise, leads to less value lost over the project cycle.

In other words, RPA helps you close the value gap between what you expected to get back from a project and the value you obtained in reality.

Document creation/processing
By automatically pulling data from various systems and inputting them into preset form fields, engineering companies can compile invoices, client emails, request-for-proposals and tax documents in a shorter amount of time. Similarly, data from invoices can be automatically processed and entered into the accounting system. If any inconsistencies are present, automatic emails can be generated and sent to clients to bring errors to their attention.

Onboarding subcontractors, suppliers and employees
Many of the mundane steps related to onboarding, including the gathering of documents and contracts, can be automated, saving time and keeping records standardized for future access.

Citec - automating document and data flow in projects
Citec currently uses RPA for automating document and data flows. Previously, due to the lack of integration between different platforms, data had to be transferred manually, which requires manpower and leaves the potential for human error to occur.

Bots at Citec are programmed to retrieve files and metadata from different systems, convert files into standardised formats and input them into different Citec systems. Since RPA can perform tasks across different applications (like a human would), data from different CAD systems can be synthesised automatically, meaning CAD-models can be kept up to date without manual intervention.

“In the document control-stage, we save 80% - 90% of the time compared to if the work was done manually”, says Citec CIO Jaako Rintala.

RPA challenges and the best practices to avoid them
Though with great potential for success, RPA is not always a guaranteed win. Vik Sohoni and Alex Edlich, Senior Partners at McKinsey & Company, say that when it comes to scaling RPA to up to thousands of bots, stakeholders (not necessarily just those from engineering companies) have found it to be a lot more costly and complicated than they originally thought.

“Several robotics programmes have been put on hold, or CIOs have flatly refused to install new bots”, Sohoni and Edlich remarked in their 2017 report. A study by Deloitte UK also found that only 3% of organisations in its study had successfully scaled RPA to 50 or more bots.

On that note, what are some major challenges that prevent the full realisation of RPA’s benefits?

Poorly mapped business processes across the whole organisation and development life cycle may kill bots on launch, as well as lead to data loss/contamination.
Changes in platforms and regulations in business processes could render deployed bots obsolete and in need of updating.
Poor expectation management - Automation is not always 1:1 with cost savings, which may lead companies to lose faith in RPA’s touted benefits before they scale and reap its full benefits.
Encountering exceptions are not planned for in advance and are not immediately dealt with by human agents, leading to substantial data corruption.
Talent management - According to Forrester Research, RPA may take away the jobs of 230 million or more knowledge workers, or 9% of the global workforce. Without proper workforce transition into roles with more value-adding tasks, you might just have an HR crisis in the making.

So, while RPA technology could be a market disruptor, it could also lead to the breakdown of business processes if not implemented properly. To prevent the latter from happening, here are some best practices relevant to not just engineering companies, but to all enterprises looking implement RPA.

Set up an RPA “centre of excellence” team whose role is to be the expert on all things RPA - from educating staff, streamlining implementation efforts and getting directly involved with client projects where RPA is involved.
Map and iron out your business processes. Sanjay Srivastava, Chief Digital Officer of Genpact explains: “Many implementations fail because design and change are poorly managed… before you implement, you must think about the operating model design. You need to map out how you expect various bots to work together.”
Standardise IT assets and bot deployment. Similar to the previous point, it is important that organisations work out which technologies in their current workflow can be streamlined/thrown out. Doing this after RPA implementation can cause nightmares in tracking down incompatible systems and bot miscommunications as a result of an already inefficient workflow.
Consider the impact RPA implementation would have on people. Disrupting the workflow of your employees may not always be a welcome move and will certainly cause some problems if HR is not included in the process. Think about how staff can be reassigned to new roles that can add more value to the company.
Start off with small successes, then scale. Deploying too many bots too fast can overload your organisation’s ability to design a governing framework that is able to control all the bots. Start with one bot, demonstrate success, and scale upwards as learn/improve your bot platform.

Finally, everyone in the company needs to embrace RPA as a long-term endeavour - and with it, an understanding of how to scale your business with an “intelligent automation ethos”. “Automation needs to get an answer - all of the ifs, thens and whats - to complete business processes faster, with better quality and at scale” says Sanjay Srivastava.

Role-Shifting and the importance of involving the IT department - how automation will change the future of engineering projects and impact its workforce
While there are not that many current cases of RPA being used in engineering, we can look at other industries and draw parallels. After all, mundane tasks such as compiling documents, replying client emails, or even maintaining compliance are not specific to engineering.

On one hand, automation cuts down manhours and enhances efficiency of repetitive tasks. On the other hand, companies who previously relied on staff to perform these repetitive tasks will find themselves with the task of either letting them go or moving them to different roles. The challenge for management will be to rethink what these higher-value roles would be for their organisation. Roles that require the flexibility and tact of human judgment, such as customer service and sales, are roles that will potentially see more manpower investment in the future.

As enterprise automation becomes more prevalent, the role of IT departments within engineering firms will grow increasingly important. IT teams should be involved early on to plan out RPA implementation in a way that doesn’t jeopardise its current legacy systems. Implementing RPA will necessitate IT departments to rethink their organisation’s current systems - what will work with RPA and what will be rendered obsolete. Due to the need for systems and workflows to be standardised, this necessary revamp of systems will no doubt be a large hurdle for large enterprises looking to leverage RPA.

All in all, the technology that enables RPA is not new, having been around for at least a decade. What makes this technology different this time round is how it has become cheaper to implement and more accessible to be wider public. Companies now offer RPA implementation services on a software-as-a-service model, with scalable subscription costs that increase/decrease depending on the number of bots you choose to activate.

Decision-makers across Hong Kong’s diverse industries - including those in engineering - must take a hard look at their own companies and recognise that automation is going to define the way enterprises are run and business is conducted in the near future. If the time is not now, then we may quickly find ourselves lagging behind the rest of the world.

• Chris Wetmore and Olga Alemany. 2019. “Transform your construction operations through RPA.” Retrieved 25 November 2019, from
• Clint Boulton. 2018. “What is RPA? A revolution in business process automation”. 3 September 2019. Retrieved 25 November 2019, from
• Xavier Lhuer. “The next acronym you need to know about: RPA (robotic process automation).” Retrieved 25 November 2019, from
• Peter Blahnik, Ernst & Young. 2017. “GNWT: Robotic Process Automation for capital projects.” Retrieved 25 November 2019, from
• Jaako Rintala. 2017. “Robotics offers great potential in engineering projects.” Retrieved 25 November 2019, from
• Clint Boulton. 2019. “RPA is poised for big business break-out.” 12 June 2019. Retrieved 25 November 2019, from
• AI Multiple. 2019. “45 RPA Case Studies: Explore RPA in your Industry & Function.” Retrieved 25 November 2019, from
• Cornelia Zou. “Hong Kong's businesses must keep up with robotics process automation.” Retrieved 25 November 2019, from
• Chris DeBrusk. 2017. “Five Robotic Process Automation Risks to Avoid.” 24 October 2017. Retrieved 25 November 2019, from
• Maruti Techlabs. 2018. “What is Robotic Process Automation & how does it compare to Traditional Automation?” Retrieved 25 November 2019, from

Poor management of resources and vendors could lead to a greater gap between expected value and the value that is delivered in reality

Automation engineers are responsible for designing, programming and testing of automated machinery and processes used to complete tasks

Home | Back About Hong Kong Engineer | Latest Issue | Past Issues
Notices to Members | Job Centre | Subscriptions | Contact Us
Terms & Conditions
Privacy Policy