• Erwan Hernot

Worst Case Scenario: Robotics Process Automation Will Strengthen Command & Control Management


Key idea: Robotics Process Automation is technically able to strengthen Taylorism. Quite a paradox when everyone is celebrating Agile and collaboration! Today management future is divided in 2 scenarios.

Since the early times, most of senior leaders were always highly aware of the input into the organization. It's exactly the same for the output. Somehow, they didn’t know the "throughput". Which is exactly why Taylorism kicked in more than a century ago. Taylorism encompassed organization's culture, organization design, systems and processes, human capability. Its main goal was to kill uncertainty (potential errors), save time and money. Says Thibault le Texier (1) taking a Marxist stance ““In the Taylorist scheme, the workers' knack is captured by the supervisory staff and, once standardized, it is reinjected into material, social and normative arrangements. The know-how housed in the minds and bodies of workers is extracted by measurement and recording, then encrypted, formalized and finally incorporated into plans, machines, production processes, training, individuals, groups, standards, stories or even a culture. ” One could add: that is the exact goal of a Robotics Processes Automation. Marx and AI: quite a story, isn’t it? Robotic Process Automation is a process of implementing technology that enables humans to configure a robot or computer software to read and analyze the current applications for performing tasks such as processing a transaction, communicating with other digital devices, triggering responses, manipulating data, etc. The companies which use workforce on a large scale basis to perform the high volume transactional process functionalities can boost their capabilities, and save time and money with the help of Robotic Process Automation software. Industrial robots are reinventing the manufacturing industry by creating higher production rates with improved quality, and the RPA robots are also revolutionizing the business process, workflow process, IT administration, remote administration, back-office works, etc. The spectrum of automation expands from simple rule-based automation to advanced cognitive and artificial intelligence automation. It could be used for rules based processes. It enables faster handling time,higher volumes and reduced errors & cost. Next level is RPA used for judgement based processes, mimicking human judgement with machine learning capability able to interpret human behavior. Next one could be a self adaptable RPA supplying predictive decisioning before, in an unforeseeable future, going towards human-like thought processing and thinking capabilities.


So, will RPA be an offspring of Taylorism and suppress managers (that it will replace) in the process ? Or a liberating tool augmenting the manager and allowing him to coach the workforce to use their brain, experiences and test & learn approach to improve customer experience? There is no established truth for now.


Scary version


“The modern, digital version of Taylorism is more powerful than we could have ever imagined, and more dehumanising than his early critics could have predicted. Technological innovations have made it increasingly easy for managers to quickly and cheaply collect, process, evaluate and act upon massive amounts of information. In our age of big data, Taylorism has spread far beyond the factory floor. The algorithmic management of the gig economy is like time cards on steroids. …Taylorism starts from the assumption that employees are innate shirkers. While there will always be some who want to game the system and put in as little effort as possible, there are plenty who don’t. When the guiding assumption of management is that employees won’t be productive unless forced to be by constant observation, it engineers low morale and pushes people to act like resources that need to be micromanaged. Too often, we become what we’re expected to be.” Brett Frischmann and Evan Selinger (The Guardian, 2017_09_25) took a look at a very pessimistic view of a very likely future… under 2 major conditions:

  1. Deep senior leaders’ assumptions about the workforce still stay the same.

  2. There is no questioning about the Agency theory (2) that aligned 30 years ago (or so) senior leaders and shareholders interests very often against those of the workforce.

In both cases, RPA aims at limiting, controlling the freedom of employees, that is to say their capacity to make choices which do not correspond to those displayed by these senior leaders. RPA will remain hidden to clear explanations and will increase pressure, regulation on the worker, hoping that these measures will lead to an improbable acceptance. Managers get back to their Taylorist role of supervisors powered by RPA. The not so subliminal message is “we know best and we don’t trust you as a manager to do the right thing, so we’re going to create a combination of processes for you that you have to follow. We’ll check you’ve followed it (because we don’t trust you), which means we’ll feel better because we know you’re doing it”. Gloomy but possible.


RPA in full Command & Control mode ?


There is no doubt that RPA will improve existing rule based processes and transform the process or the function, creating a new product, service or business using AI at its core. But RPA can’t catch the extraordinary complexity of human organisations for now and for a foreseeable future. Nevertheless, many people assume that change can be harnessed by implementing processes. But I believe a focus on process is a too narrow approach to managing change. We need continuous business improvement, not just process improvement. From a complexity perspective uncertainty cannot be dealt with by (only) implementing processes be they powered by RPA. Uncertainty applies to the whole system including the execution of any processes. But how can one improve a system in which everything is unpredictable, including the processes? Complex problems are those that behave unpredictably. Not only are these problems unpredictable, but even the ways in which they will prove unpredictable are impossible to predict. So, could Robotics Process Automation be fit for the job? The ability of this tool/solution to automate depends on three factors: the type of input it can read; the amount data it can process; and the nature of output it can generate. How would you feed the machine with politics, power and influence struggles that exist in all human organizations? Enterprises can supercharge their automation efforts by injecting RPA with cognitive technologies such as machine learning, speech recognition, and natural language processing, automating higher-order tasks that in the past required the perceptual and judgment capabilities of humans. The wrong assumption behind this “RPA as structure on steroids”, is that reality will adapt to its design. RPA is indeed built on models. This use of models (of behavior, organization etc.) replaces, for the machine, the knowledge of reality. The solutions RPA could provide are never based on actual knowledge of the problem but on reductive assumptions. Besides this complexity not fully managed by the RPA, exemplary business process management performance demands that leaders work collaboratively and cross-functionally. They don’t. How can you make a Bot understand that the CFO hates the CIO and that each one is engaged in a turf war disturbing the rationality on which its algorithms are based? Effective managers don’t rely only on formal authority as their source of power. And RPA would be formal authority by definition : there is nothing less formal than an algorithm. RPA may get control over people but won’t produce commitment in the process: people are not engaged and inspired when controlled by old fashioned managers. So control by RPA won’t go very far in terms of continuous efficiency. Commitment comes with empowerment or sharing influence. RPA does not understand that there are so many sources of power, and that managers’ responsibilities include growing their own power and influence in order to manage key interdependencies and relationships in the company. RPA would be at pain (if you can say so for a machine;) with conflicts based on situations not seen before. To eliminate political conflicts, RPA would have to eliminate diversity, including the division of labor and specialization of employees. The greater the task specialization in the organization, the more likely there will be disagreements. But specialization is often necessary for organizational efficiency and productivity. Diversity and conflict are also essential ingredients for creativity and innovation. The manager is responsible for managing the inevitable trade-offs and for equitably negotiating and integrating his or her team’s interests with those of others, like superiors or peers in other functional or geographic areas. Rather than engage in political infighting, effective managers seek creative win-win solutions, and avoid participating in zero-sum negotiation behavior. Quite a task for a machine! A Robotics Process Automation in full Command & Control mode is not happening now. When RPA is able to mimic human intelligence and acquire human-like thought processing and thinking capabilities, we’ll think again though…


(1) Thibault Le Texier "Le maniement des hommes" 2018

https://www.theguardian.com/commentisfree/2017/sep/25/robots-taken-over-work-jobs-economy

(2) Agency theory : https://www.investopedia.com/terms/a/agencytheory.asp



Les techniques managériales ne sont pas neutres...

  • Noir LinkedIn Icône
  • Noir Twitter Icon