How robots are changing the face of banking
Robotics, enabled by artificial intelligence and machine learning, is proving to be a game changer that can bring unique operational efficiencies to the financial services industry.
- According to CB Insights, investment into robotics rose by 115% in 2015
- Robotics process automation can bring notable time and cost efficiency improve productivity and operational improvements to financial institutions
- The application of robotics in financial services is gathering pace but most banks are still in the early stages of adoption
Move aside people, robots are here! Robots are becoming smarter, faster and cheaper and their deployment across industrial usage such as automotive and electronics has surged in the last couple of years. From industrial robots to self-driving cars, robotics technology is becoming more cognitive-enabled that replicates human skills and intelligence.
For the past years, investment into the robotics sector has risen rapidly. According to CB Insights, the number of funding deals in robotics globally nearly doubled from around $273 million in 2014 to $587 million in 2015. The investment growth in 2015 was 115%, as compared to 55% in 2014 (Figure 1).
Investments into robotics is rapidly rising
The International Federation of Robotics (IFR) estimates that there will be a total supply of 1.4 million industrial robots between 2016 and 2019, forecasting an average growth rate of 13% in the supply of robots from 2017 to 2019. In 2015, the worldwide annual supply of industrial robots increased by 254,000, a 15% growth from the previous year.
While robotics hardware has found increasing industrial applications over the last few years, its entry into the services sector, especially financial services, is quite recent. This technology is now touted to bring fundamental changes to the way banks operate heralding a new era in self-service banking. For example, millenials have high digital expectations from their banks. They are likely to be more willing to engage with robots for financial transactions and ready to trust them for complex activities.
To meet these expectations, banks are using robots in multiple processes. There is what we call Software robotics process automation (RPA), which is becoming a game changer in back-end computing. Humanoid robots are providing customer service at select branches, bringing a unique ‘wow’ element to clients. Likewise, robo-advisors are being deployed in areas like investment advisory.
Robots for novel customer service
Robots come with unique advantages – they are time and cost efficient, improve productivity, deliver superior results, and can work without rest over repetitive tasks. When enabled with cognitive computing, artificial intelligence (AI), and machine learning capabilities, robots can be trained to operate autonomously. They can also learn how to improve performance and accuracy with little or no human input. In addition, multi-lingual language processing and voice recognition capabilities allow robots to interact and conduct seemingly intelligent conversations with customers.
For example, Bank of Tokyo- Mitsubishi UFJ (MUFG) introduced Nao, a 58-centimetre (1ft 11)-tall, 5.4 kg robot developed by Aldebaran Robotics – a France-based subsidiary of Japanese telecom and internet giant SoftBank. It is equipped with a camera and microphone and has visual recognition and remote control capabilities. It can recognise 19 spoken languages, interact and communicate with customers in branches, and provide response to queries.
SoftBank also developed Pepper, which is being used by Mizuho Financial Group and Emirates NBD. It has an interactive tablet that helps augment communication with users. The robot will entertain customers with games and multimedia functions, while providing basic product information. And banks intend to continue conducting researches that will help them expand the cognitive capabilities of these robots to make them more useful and efficient.
Meanwhile, chatbots, which are AI-enabled virtual assistants that communicate via text rather than speech, are gaining traction in the market. Hoping to cut costs and improve service, banks such as Bank of America and Royal Bank of Scotland are planning to deploy chatbots to answer customer queries that mimics human-like interaction.
“Cognitive systems can interact with the natural language, like having conversations, to answer questions. A robot can go out into the database and content management system, haul and interpret documents, and then use them for practical purposes. It can understand the meaning and more importantly, the context and the intent of the conversation. It can apply insights and information in a very practical way and finally it learns from experience,” said Vincent Kasten, executive partner for cognitive solutions at IBM, during the Misys Connect forum in Singapore.
Robotics process automation (RPA) brings unique efficiencies in financial services
The need to introduce new cost efficiencies, improve digitisation and faster transactions have forced banks to rapidly explore the use of robotics in processes. RPA, as defined by the Institute of Robotics Process Automation (IRPA) allows company employees to configure computer software or a ‘robot’ to capture and interpret existing applications to process transactions, manipulate data, trigger responses, and communicate with other digital systems. This requires a system to process vast amount of information, use algorithm to detect patterns, and enable rules-based decision making.
RPA, at its core, is designed to address the monotonous and tedious repetitive tasks that can be handled quickly and efficiently with technology that mimics humans, without their intervention. It bridges legacy systems to streamline data management and processes. Moreover, it does not require IT architecture changes and deployment can be easily scaled as needed.
“RPA can be implemented across the board to all types of processes, except probably the ones that require a high level of cognitive capability. It is a journey that requires creating a dedicated centre of excellence, training people, managing process and data security, and building operating model around RPA. The automated process can then be replicated to deliver savings week after week. If this is done ‘right’ then an institution can start with 40-60% cost reduction within the first year and extend up to 70% cost savings,” said Daniel Dines, CEO and founder of RPA software company UiPath
“Right now, collectively as an industry, the penetration of RPA across Fortune 1000 companies is about 40% in different phases of adoption, but if we look at the pipeline, it is likely to grow to 70%-80% in the next couple of years,” he added.
And today, the application of RPA is rapidly scaling. Grand View Research estimates that the current global market for RPA is worth $125.2 million, which it expects to grow to $8.75 billion by 2024.In addition, the adoption of RPA in the banking and financial services sector is expected to grow at a CAGR of over 65%.
One of the key element in the development of robotics is AI, which enables computers to think and analyse information like a human. “RPA is the eyes and hands while AI is the brain. They both need to coexist, with RPA becoming a key enabler for AI. RPA is largely focused on rules based processes and brings discipline into how one thinks of processes and reengineers them. With RPA one can collect enough data to create machine learning and cognitive process,” Dines explained.
Robotics is expected to have a lasting impact on improving efficiencies. KPMG predicts that RPA can cut cost for financial services firms by 75%, while IRPA believes it can result in 25% to 50% cost savings. KPMG adds that in the next 15 years, up to 75% of existing offshore jobs will be performed by robots. IRPA believes that a software robot can cost as little as one-third the price of an offshore full-time employee (FTE) or one-fifth the price of an onshore FTE.
Besides the cost savings, efficiency improvements through higher productivity, ability to work 24x7, and greater accuracy by reducing human errors can be achieved. Their ability to collect and mine vast data and complete audit ability are especially useful in areas like compliance and regulatory reporting. Furthermore, these robots can be deployed and scaled with ease and agility (Figure 2).
RPA enables new efficiencies and cost savings
Source: Asian Banker Research.
Rising application across financial sector
High volume, manually intensive, and prone to risks and human errors processes are prime candidates for RPA.
For example, ANZ is using RPA in processing payroll, account payable, mortgage procession, and human resource (HR) functions. ICICI Bank, meanwhile,uses RPA to perform over one million banking transactions in back-end operations per day, reducing response time by 60% and improving accuracy. These software robots are deployed in over 200 business process functions of the bank across retail banking, agri-banking, trade and forex, treasury, and HR. These processes include addressing change requests, ATM query resolution, and verifying know-your-customer compliance.
Likewise, Barclays Bank implements RPA across a wide range of processes such as fraud detection, risk monitoring, account receivables processing, and loan application. Piraeus Bank Romania has implemented a RPA solution from UiPath in assessing retail credit and preventing fraud by connecting with 15 applications. The typically long and arduous process (retail credit fraud prevention) that takes around 45 minutes was automated and reduced to only around 20 minutes.
Robo-advisory, an algorithm-based intelligent and automated investment solutions, is another application of robots that has been gaining increased attention. Robotics-based investment advisory can reach an average investor with ease while raising fee income for the institutions. KPMG estimates that by 2020, robo-advisors will manage $2.2 trillion in the USA. For instance, robo-advisor company Betterment is reportedly managing over $5 billion assets under management.
Intelligent analytics and machine learning are rapidly being applied in robo-advisory to assist in decision making. “We believe that the robotics and machines learning is going to be something to be improved and the biggest area in robo-advisory is dynamic risks assessments,” explained Ned Phillips, founder and CEO of robo-advisor company Bambu.
“Finance will have to move towards machine learning but this will not be sudden – there will be a period of change,” he added.
Banks need to scale adoption
Many banks are still in the early stages of adopting robotics and are yet to scale it across different processes, facing a number of challenges.
One challenge is that there is an upfront capital investment that needs to be justified and there is a need for a champion or owner to promote it within the organisation. Additionally, automating operations will likely have an impact on staff requirement. Inevitably, roles made redundant by automation and robotics will need to be reconfigured and staff will need to be retrained, reskilled, and redeployed, which will require an intense change in the management.
Aside from this, as institutions expand their use of digitally connected robot workforce, exposure to potential cybersecurity threats also increases. As robotics merge with new technologies and get interconnected, for example, through the internet of things (IoT), security lapses or potential hacks might increase. A clearer robotics automation strategy and framework will assist institutions to better manage the threat.
“The biggest bottlenecks we see today are –lack of trained resources capable of delivering RPA, an internal inertia that requires a change management process, and heavy scrutiny and concerns by the institutions with regards to data security. There is also a perception the robots will lead to staff attrition but the truth is that robots replace humans in dull processes and empower the same human to do high value processes. However, this message needs to be well articulated by the key stakeholders and sent across the institution,” Dines explained.
What is next?
Clearly, the role played by robots is increasing and expected to gather pace, although RPA is still in the early adoption phase in the financial services sector. Yet, this technology is expected to evolve and scale in the next years, although there are still existing gaps between humans and robots, which will continue to drive future developments. The true benefits of RPA can achieved through greater deployment towards end-to-end automation in banks. They should first implement process reengineering to change the inefficient processes and workflows before applying RPA.
Nevertheless, banks should not just focus on cost efficiencies. They must also improve their competitiveness by ensuring that they enable higher accuracy, reliability, and better service capability, while up-scaling their cognitive capability through AI, deep learning and robotics.
Keywords: Robots, AI, RPA, MUFG, Mizuho, UiPath, ANZ, IoT, Technology, Robo-advisors