Myriad of possibilities with Robotic Process Automation
By Ian Herbert November 4, 2016
- Robotic process automation (RPA) is rapidly becoming 'business as usual' in a wide range of service sectors
- There is a considerable overlap between the concepts of automation and robotics
Predicting the future in any detail is difficult. For example, it has been assumed traditionally that low level, repetitive jobs will be automated and unemployment will start from the bottom of the labour market. Instead of ‘mundane’ low-level jobs that are being targeted for robotics, it may be more cognitively demanding, middle-level jobs that are reconfigured for robotics.
Once the domain of science fiction and then advanced manufacturing, robotic process automation (RPA) is rapidly becoming ‘business as usual’ in a wide range of service sectors from health care to transport and logistics. Not surprisingly there is a lot of hype, both in terms of our relationship with ‘humanoid’ machines and the likely extent of job losses. The main assumption being that it will be the low-level jobs that will disappear first.
Differentiation between Automation and Robotics
There is a considerable overlap between the concepts of automation and robotics. One way of distinguishing between the two is to see automation as a largely technical capability, focused on replacing human mechanical actions. On the other hand, robots can tackle relatively cognitive tasks which require the capability to sense the surrounding environment and react flexibly towards an overall outcome.
Both automation and robotics consists a combination of software and hardware. Satellite navigation is principally about the processing of data from four sources; the motorist, the preloaded map data, the satellite positioning system and the environment ahead of the journey.
By contrast, the type of ‘robots’ typical in car factories are principally hardware devices that have been programmed to carry out a relatively limited series of operations on a certain model or model(s) of vehicle. We tend to call these mechanical arms ‘robots’ because we can see them and they tend to very hugely expensive, but they are more akin to advanced automation, whereas in-car navigation systems, whilst cheap and portable, are more accurately robotic as they can make sense of a dynamic environment and interact with humans as appropriate.
RPA needs to be seen as a part of longer term journey towards lights-out processing, which is first enabled by total digitalisation, sensible self-service systems and appropriate standardisation - all of which may create the possibilities for automation to: eliminate manual operations, enhance present operations, augment information flows and management decision making capability, and/or provide further options for robotic management and decision making. The key here is ‘augmentation’ not ‘replacement’ of human labour.
The future is now
Automation systems through big data and analytics tools are already in position to aid the middle managers with ad-hoc report generation. While this activity would be done traditionally by experienced managers, in a digital world driven by internet self-service, most of these decisions is now performed by robotic software that learns from past experience and interacts with wider information sources on a 24/7/365 basis.
These lights-outs processes can handle all aspects of routine customer administration and set prices dynamically for air/bus/rail tickets and for online retailers like Alibaba, Amazon, eBay etc. On-line retailers take the process a stage further by employing sophisticated algorithms which automatically serve up appropriate suggestions to customers. Such automation tools, mimic the actions of a ‘human store retailer’ when a customer comes in, evaluates the customer’s background and stated preferences and then make sensible suggestions on what products would that particular customer might be interested, noting any opportunities for upselling.
The best example of a robot is the ‘driverless’ vehicle. Once programmed with a destination a car can drive itself by sensing its immediate road environment (including other cars) in conjunction with other data such as satellite data on road and travel conditions. Such cars will also have the ability to learn from their environment and thus, optimise regular journeys in co-operation with other driverless cars.
The obvious advantage is that it frees driver time for thinking and working there are other important advantages that will create new business models around driving. In a co-ordinated planning system, a number of cars might drive themselves back to popular pickup places at different times during the day to optimise traffic flows, e.g. to MRT stations in the morning and back to workplace areas in the afternoon. Singapore, for example, will soon have a self-driving taxi service operated in the city-state; where selected members of the public can hail for the taxi through their smartphones.
The point we are making is that automating routine tasks and applying robotic technology to more cognitive or less routine tasks has limitations if only in terms of cost versus the benefits, however, there are new possibilities for new ways of human working and new business models if adaptive changes are made in the operating environment. The big wins will occur as robotics and automation are combined.
Ian Herbert is the deputy director of the Centre for Global Sourcing and Services at the School of Business and Economics, Loughborough University.