Categories

The Future of Work in Oil and Gas

It’s not just robots applying for your jobs, writes Dominika Warchol Hann, in this week’s post.


A shrinking share of blue collar jobs and the increasing reliance on foreign outsourcing are driving political debate and making blue-collar-dependent companies take note. As you’ll soon find out, the prognosis is for these trends to not only increase, but to accelerate due to the rise new ways of working. These new ways of working include the increasing prevalence of automation, crowdsourcing, onshore outsourcing and offshore outsourcing.

Many jobs can be outsourced to something or someone other than your staff. The remainder will be disrupted by new ways of working


Regardless of whether a machine, another person, or a crowd performs the work, ultimately we’re asking ourselves the same question that organizations looking to emphasize their focus on core competencies have been asking for decades:

What can we conceivably outsource? How can we outsource in the most effective way (in terms of both speed and cost)? How can we maintain quality while doing so?


While no definitive numbers are available to tell us what the future holds, a number of papers published suggest the following:

  • Numerous jobs across manufacturing to professional services can be either replaced or highly augmented by automation and / or outsourcing (McKinsey suggests that 45% of all activities could be partially or fully automated, Deloitte UK suggests that between 25% and 31% of jobs in Business Services are at high risk of automation)
  • The rise of creative and complex crowdsourcing is filling the gap between outsourcing and automation. Up to 40% of US workers are predicted to be contractors by 2020. This amount may rise with additional crowdsourcing and freelancing platforms of the shared economy
  • For work that must be done in-house, but not necessarily in the office, the continued growth of remote work is changing organizational real-estate requirements

The bottom line: the way staff do work today is not the way most will be doing work tomorrow.

Explainable, repeatable, modular, and technology-enabled tasks are at greatest risk of disrupting “who does the work”

Predictable (explicit and codifiable) tasks are ripe for disruption


If you can explain a task to a human, you can often explain it to a machine. For decades, tedious manual processes have been outsourced to a machine. In recent years, with the advent of increasingly sophisticated robotic process automation (RPA) technologies, more decision-based tasks and analyses are being earmarked for machines.


At a recent conference, Xerox and their partners demonstrated their sophisticated RPA capabilities. Forms and written documents were digitized, ingested by an artificial intelligence, machine and processed according to rules that an organization required (consider: field tickets, invoices, and purchase orders). Xerox is not the only one. RPA services and consulting are provided by a plethora of professional services firms and technology companies (including Deloitte, which – full disclosure – is my employer).


In the world of creative and analytic-heavy tasks, Orbital, a satellite-imagery analytics company, has been training artificial intelligence to track the volume of crude oil currently being transported on the ocean. By analyzing satellite imagery, their AI considers a ship’s shadows to determine a ship’s buoyancy and load. In each case, the work being replaced is that of a researcher or a human analyst. This is just one type of explicit analysis that a machine learning algorithm could potentially disrupt.

Modular tasks are also positioned for outsourcing to the crowd


Modular tasks are self-contained activities with no or few dependencies on other activities. The more modular and less sequential a task is, the more likely it is be disrupted not only by automation and traditional outsourcing, but also by crowdsourcing. Crowdsourcing models differ and include simple crowdsourcing, creative crowdsourcing and complex crowdsourcing. At its root, crowdsourcing requires that a specific task or problem be set before a large group of individuals. The individuals then self-organize around solving the problem and multiple potential solutions arise, of which the best is ultimately adopted by the sponsor.


In some ways, crowdsourcing acts as an intermediate between contracting and offshoring and automation. Crowdsourcing can undertake simple automatable tasks or develop complex innovations (e.g., the X-Prize) more cost effectively than automation, outsourcing, or undertaking activities in-house.


O’Neil et. al. (2013) use the example of medical record coding. As the task does not need to be completed sequentially, it is easy to modularize. By creating an effective work-queuing and reward system, their proof of concept effectively reduced costs and competed directly in quality and with automation and offshoring.

Contraindications to outsourcing, crowdsourcing and automation are tacit knowledge, risk and adoption by the public


Tacit knowledge comprises knowledge that is integral to the functioning of an organization but is non-codifiable. It often enables workers to react in the right way during an unpredictable situation. For example, tacit knowledge may comprise of situations where an email’s tone “does not feel right” or, where the statement “he set me on fire” is followed by the dispatch of an ambulance and police, not a fire truck.


Risk. British Airlines’ recent system crash has been, at least partially, placed on the shoulders of its outsourced operations. Other operations cannot legally outsource certain tasks (e.g., 911 call-takers).

“The bizarre thing about neural networks…You cannot tell what they are picking up. They are like black boxes whose inner workings are mysterious.”

Sebastian Thrun


For automation, certain decisions made by AIs might never be sufficiently explainable to be trustworthy. That is certainly the perspective of Andy Maloney, owner of FORident Software, creators of HemoSpat, a bloodstain pattern analysis software that enables police investigators to analyze crime scenes. “While you could use machine learning algorithms to analyze and classify some types of bloodstain patterns, when a detective testifies in court they need to be able to show how a specific solution was arrived at. Often we don’t know what features these algorithms are choosing to make their decisions, so they could be biased.”

Consequently, individual organizational risk must be taken into the decision to outsource, crowdsource or automate.


Adoption by the public. Diagnostic and research services have been well documented as ripe for disruption. Not only can computers do them more cheaply, they can often do them better (e.g, completing diagnostics in radiology). So why aren’t we seeing more adoption of these technologies? In part, it is a lack of faith. Even if a computer is making a decision, having a human available to provide direction and a second sober look is key.

What does it all mean?


The rise of automation, the continued growth of outsourcing, and the new social-economy ways of working (crowdsourcing) are shifting the future of work. They will likely culminate in an exaggeration of the trends faced by the US due to the rise of offshoring, as identified by economist David Autor:

  • The number of workers that organizations will need will likely drop, first in blue collar jobs and then in white collar
  • The wages of employees in adjacent professions, industries and levels will also drop, creating cost savings opportunities for organizations (the broader ethical and societal implications are another matter)
  • As additional strain is put on organizations to be cost effective, R&D budgets may be the first to be cut. Consequently, it will be crucial to determine what an organization’s strategic differentiators are and how they can be maintained in an increasingly cost-competitive world

Consequently, when looking at the outsourcing, crowdsourcing or automation decision it is important to not only address the “could you?” but also the “should you?”. So where might an organization begin such an assessment?

The greatest benefits of the future of work are highly dependent on the industry in question. For commodities, a key workforce strategy focuses on cost reduction


Labour is one of the most significant costs to an oil and gas organization. As such, the opportunity to increase labour productivity, reduce labour costs, and address labour-related real-estate costs is key to delivering greater value for your shareholders.

As AI systems, robotics, and cognitive tools grow in sophistication, almost every job is being reinvented, creating what many call the “augmented workforce.” As this trend gathers speed, organizations must reconsider how they design jobs, organize work, and plan for future growth.

Deloitte, Human Capital Trends 2017


A strategic analysis of the future of work would consider:

  • How your organization differentiates itself and the key services that create value for your shareholders; then identify the core tasks that drive those services
  • The opportunity for outsourcing, crowdsourcing or automation against the criticality of these core value-creating tasks; the strategic “should we?” countering the mathematical “could we?” argument (considering arguments around tacit knowledge and risk)
  • An analysis of automation and outsourcing against non-critical tasks, with a greater focus on cost optimization. While this analysis would still consider risk and tacit knowledge, these tasks are non-core to the value creation of the organization and are therefore not as likely to pose an organizational risk should they be outsourced, crowdsourced or automated
  • Rethinking the way tasks are performed: whether they are done remotely, by loosely knit teams, or in a more structured hierarchy, and identifying people-tasks that can be further augmented through automation (e.g., RPA, advanced analytics) to create greater efficiencies and insights
  • Linking remaining tasks to capabilities and generating new roles, which are then consolidated into new organizational structures organized around key organizational value drivers

Regardless of the method of workforce analysis undertaken, there is a great deal of opportunity for your organization to redesign how work is done, enable a value-creating workforce, and rethink how value is generated for your stakeholders and employees.

This article was written by my colleague Dominika Warchol Hann, who also drew the artwork.

Share this ... Facebooktwittergoogle_plusredditlinkedinmail
No Comments

Leave a Reply

%d bloggers like this: