Drones, mechanical robots, and robotic software processes are augmenting, assisting, and, in some cases, replacing human workers in almost every industry. The common assumption that drones and robots threaten primarily manual labor is not entirely correct; robotic disruption may also soon affect areas of white-collar work. The authors discuss the possible ramifications of automation in the accounting profession.
* * *
As with many professions, accounting and auditing may in the near future be disrupted by recent innovations in remote-controlled drones, mechanical robotics, and robotic process automation (Daniela Rus, “The Robots are Coming,” Foreign Affairs, July/August 2015, http://fam.ag/2rUAEG9). Computers have already become ubiquitous in business, and in the future drones and robots of all kinds will as well. In short, the profession stands on the cusp of a technological sea change: the “dronnovation” of accounting.
Dronnovation, as the authors conceive it, applies to three different kinds of technology: drones, mechanical robots, and robotic processes (or bots). The goal is not to replace human accountants, but to find ways that machines can assist and support humans (Thomas H. Davenport and Julia Kirby, “The Knowledge Jobs Most Likely to Be Automated,” Harvard Business Review, June 23, 2016, http://bit.ly/2rQZBTN; Davenport and Kirby, “Just How Smart Are Smart Machines?” MIT Sloan Review, Mar. 15, 2016, http://bit.ly/2rRr4og). Drones are better at inspecting, observing, and monitoring than humans, and robots are better at lifting heavy objects, and working around the clock (Rus 2016; Tom Simonite, “Mining 24 Hours a Day with Robots,” MIT Technology Review, Dec. 28, 2016, http://bit.ly/2qgBqNs). In the realm of bots, automated routine data collection procedures and data analysis applications will converge in a cloud-based, real-time monitoring and auditing system supervised and utilized by accountants, who will be able to focus on more complex and judgmental tasks.
Declining cost is making devices such as drones and robots more attractive for automated applications (“The Robots are Coming,” Deloitte UK, 2015, http://bit.ly/2rAwkiJ). Drones are more flexible than humans and can carry various data recording and generating devices. Miniature drones the size of large insects can fly between closely stacked warehouse pallets (Steve Banker, “Drones and Robots in the Warehouse,” Forbes, June 24, 2016, http://bit.ly/2qQbpap) or into narrow, hard-to-reach spots, allowing them to collect information that may have been impossible to gather previously (Deloitte UK 2015).
Drones are frequently used for photography, surveillance, observations, and recording in many industries such as agriculture, filmmaking, mining, and surveillance. With cameras and sensors attached, drones can record video, map, and detect. These readings and videos can then be fed into software applications for understanding and analysis. Such use of drone technology could easily be applied to asset valuation, observation, and inventory counts.
The surge in drone use is partly due to recent Federal Aviation Administration (FAA) actions. On August 29, 2016, the FAA relaxed restrictions on personal/hobby drone use and defined procedures for commercial use (“The FAA’s New Drone Rules Are Effective Today,” FAA Website, Aug. 29, 2016, http://bit.ly/2rjpXQj). Drones used for business purposes must be navigated within sight unless a special permit is granted for unobserved flight. Furthermore, for outdoor drone use, the pilot must know local ordinances and restrictions, and any flight over 400 feet above ground also requires special FAA permission. Drone use on private property is less restrictive, if outdoor flights stay below 400 feet.
Robots are being used to work alongside human workers for many inventory projects, which could translate to information collection in accounting and auditing.
Robots are being used to work alongside human workers for many inventory and warehouse projects (Patrick Clark and Kim Bhasin, “Amazon’s Robot War is Spreading,” Bloomberg Technology, Apr. 10, 2017, https://bloom.bg/2rAWEtk), all of which could translate to information collection in accounting and auditing. Likewise, robots could reduce the danger, monotony, and drudgery of certain tasks. Robots, and by extension drones, are best suited for predictable tasks where the cost per error is low (Dhar 2017) and the risk is well managed. Some industries have connected their robots to a cloud-based alert system that analyzes the data feeds and prevents disruptions to manufacturing by constantly monitoring for potential breakdowns (Thomas Black, “GM Hooking 30,000 Robots to Internet to Keep Factories Humming,” Bloomberg Markets, Apr. 4, 2017, https://bloom.bg/2rV4Vol).
Bots are already being utilized by many industries. According to Deloitte’s 2015 survey of global business leaders, automation is a top software priority (Deloitte UK 2015). Executives anticipate that automation could deliver value for clerical and spreadsheet processes that are repetitive and rules based; many transactional processes, data manipulation, and data analysis tasks could also be automated. According to the Deloitte survey, 91% of executives anticipate that accounts payable transactions will be automated, followed by 55% for travel and expense, 36% for fixed assets, 27% for general ledger, 18% for financial reporting, and 9% for payroll and payment cards. In fact, many anticipated areas for future automation exist in accounting, where robotic processes could reduce costs and improve the quality of routine tasks. In this paradigm, accountants are envisioned as managing and monitoring the bots and personally examining outlier events and higher-risk situations. This will allow accountants to focus on those transactions that require nuanced human analysis and not worry about routine high-volume transactions. Drones and robots could be the accounting digital assistants of the near future, completing the more predictable and mundane accounting tasks, both physical and mental.
Three Steps towards the Automation of Accounting
Research and practice suggest that automation of a manual process should be undertaken incrementally (Michael G. Alles, Alexander Kogan, and Miklos A. Vasarhelyi, “Audit Automation for Implementing Continuous Auditing: Principles and Problems,” working paper, 2008). The optimal process has been suggested as augmentation: “Starting with what humans do today and figuring out how that work could be deepened rather than diminished by a greater use of machines” (Thomas H. Davenport and Julia Kirby, “Beyond Automation,” Harvard Business Review, June 2015, http://bit.ly/2rhOdmC). This integration approach to automation can already be observed with personal income tax filing (Vasant Dhar, “Robots Will Soon Do Your Taxes. Bye-bye, Accounting Jobs,” Wired, Feb. 2, 2017, http://bit.ly/2qQf4ot).
Fifty years ago, preprinted tax returns were filled in manually by typewriter after the numbers were totaled by hand or calculator; over time, spreadsheets were used for calculations, software programs offered report templates, and eventually even filing became electronic. As technology replaced more of the manual tasks for taxpayers and professional preparers, tax preparation services came to be hired mainly for more complex returns. Now less complex returns can be handled by bots; in fact, a major tax filing service provider recently announced that it would be preparing and filing returns with an AI application. Less predictable personal and business returns, however, still require nuanced human judgment (Davenport and Kirby 2015).
As seen with tax services, the integration and eventual automation of manual and routine thought processes in a profession entails several steps:
- Integration of manually directed automation tools to assist the work task
- Augmentation of the work task with semiautomated tools
- Work task is heavily automated, with human monitoring and analysis of the resulting big data. (Davenport and Kirby 2015)
The accounting profession, with its requirements for inventory validation, transaction and event monitoring, and asset evaluation, stands on the brink of its own automation revolution. Once the profession begins utilizing drones, robots, and bots, the possibilities are endless. This process will comprise three steps—adoption as an extension of the accountant, the augmentation of accounting tasks, and finally, the full automation of accounting.
Adoption as an extension of the accountant.
In this step, for example, an accountant would either pilot (or observes piloting) drones for tasks that are either routine, laborious, or dangerous. This includes calculation and verification of inventory, asset observation and verification, and surveillance. Except for collision avoidance and energy management, the drone is manually dependent on the pilot for flight patterns and objectives. Even in this stage, however, the drone can greatly enhance the observations of the accountant/pilot due to onboard sensors and cameras; these readings can be viewed in real time and recorded for later analysis. Recent technology allows drones to read 20 feet deep through boxes, cartons, and crates in all directions above or below its location (conversation with SmartX Technology by the authors).
Drones could also be used as flying cameras (e.g., conducting surveillance of property and other assets, collecting video feeds and sensor readings) to ease the process of information gathering. Several local municipalities now use drones to inspect bridge structures, gathering highly detailed information about their conditions in a fraction of the time previously required with humans (“Drones and Bridge Inspections—Changing the Process,” RDO Integrated Controls blog, Oct. 10, 2016, http://bit.ly/2rUVfKs).
For accountants familiar with video games, controlling a drone or a robot could feel natural.
The toolbox of the future accountant could very well include a robot and a drone, as these are technologies that can greatly ease the task and improve the quality of information collection. A robot on the ground could navigate to the correct warehouse location (Clark and Bhasin 2017); the accountant could then pilot a drone to assess/count inventory at upper levels of the pallet rack. This toolbox could also include smart glasses that allow the pilot to see in real time what the drone is seeing and recording.
For accountants familiar with video games, controlling a drone or a robot could feel natural. According to recent users of robots in warehouses, “working with robots is a fun thing to do” (Clark and Bhasin 2017). Alternatively, accountants could wear special glasses that pilot the drone based on head movement and eye tracking (Fintan Corrigan, “FPV Goggles for Drones and Experience the Thrill Of Flying,” DronZon, May 1, 2017, http://bit.ly/2rAAwPy).
Augmentation of accounting tasks.
In the second phase, drones, robots, and bots are programmed to automatically conduct routine and predictable tasks and augment an accountant’s ability to complete tasks. An accountant configures their parameters but minimally controls their navigation or operation. Once configured for a task, the machines can proceed with a high degree of self-sufficiency until the task is completed; an accountant monitors the real-time data feeds and adjusts their tasks as needed. Much of the data collected is integrated into the company’s cloud-based system with other systems and data feeds. These data feeds are not continuous, but rather streamed only when the accountant uses the drone or robot with an objective. Because of the sporadic nature of this data collection, analysis is still batch oriented, but more predictive and prescriptive in nature due to the volume and variety of data available. Drones, robots, and bots at this stage contribute to an objective of near-continuous monitoring and auditing.
Here the automation of drones, robots, and bots converges to create a fully autonomous application that can make its own decisions and conduct many, if not most, physical and routine data tasks in accounting. Bots automate manual processes such as data entry, data cleaning and preparation, data transfers and combinations, and transaction entries and linking. Bots are now beginning to automate data analysis and exhibit greater intelligence (Davenport and Kirby, March 2016). Drones and robots, meanwhile, are demonstrating great efficiency and accuracy with physical tasks such as inspections, observations, and surveillance. These three technologies can be interconnected within a cloud-based platform and communicate seamlessly (Jyrki Karasvirta, “Cloud4Drones Platform Enables Live Use of Drone-Collected Data,” EIT Digital press release, Apr. 18, 2017, http://bit.ly/2qgX3x2).
The human accountant in a “dronnovated” environment will monitor the bots, robots, and drones, tweaking their algorithms, maintaining their security, and intervening when necessary.
At this stage, drones, bots, and robots can operate continuously, conducting multiple tasks as necessary. They are connected on a real-time continuous basis to a cloud-based platform with many integrated software applications and feedback loops (Karasvirta 2017). Much of the data is collected automatically and is accessible in a cloud-based centralized location. These data feeds are integrated with other pertinent information by bots (Davenport and Kirby, March 2016). Data integration can be a lengthy, time-consuming process when conducted by humans but a fast, efficient process when done by bots (Deloitte UK 2015), which can use workflow and business rules technology to interact with multiple data sources and systems. Accountants will control these various platforms and systems by modifying and updating the rules and analyzing exceptions.
What about the Human Accountant?
The human accountant in a “dronnovated” environment will monitor the bots, robots, and drones, tweaking their algorithms, maintaining their security, and intervening when necessary. Accountants will also analyze any anomalies that occur, focusing more on the transactions that require nuanced in-depth analysis, forensic investigation, and fraud examination. This will require “exercising professional judgment, having an independent mind-set, maintaining objectivity, and other baseline financial understandings” (Helen Brand, quoted in Kristin Kang, “Robots Could Soon Take over Parts of the Finance Function,” CFO Innovation, May 4, 2016, http://bit.ly/2rB7Dmi). Even if the robots and drones are incorporated with artificial intelligence applications, there are decisions and analysis that require intuition and softer skills that robots have yet to demonstrate (Michael Cohn, “Deloitte Leverages AI and Analytics for Audits,” Accounting Today, Nov. 14, 2016, http://bit.ly/2qh8vJ5). In addition, auditing and accounting standards likely will need to be modified to provide guidance as to best practices regarding information assurance, a task that can only be done by humans.
Automation has already impacted the financial industry, bringing operational efficiency, accuracy, and standardization to both process-oriented and analytical tasks (Kang 2016; Penny Crosman, “All the Ways AI Will Slash Wall Street Jobs,” American Banker, Mar. 17, 2017, http://bit.ly/2qXhsu0). Financial companies will spend more than $1.5 billion this year on automation technologies, and this amount is expected to grow significantly. It is doubtful that the accounting profession can remain unaffected by this trend, as accountants must work with many industries that are undergoing automation. To navigate in this increasingly technically complex environment, accountants will need to avail themselves of drones, robots, and bots. The time has come for the dronnovation of accounting.