CDN NEWS | 
US NEWS 
By Geoffrey Cann
A large integrated oil and gas company has invited me to address their global center for data excellence. That such an organizational creation is necessary is both telling for some and enlightening for others.
Bus ta Move
Many years ago, I helped a bus operator that provided worker transportation to oil and gas facilities with a huge technology upgrade. Part of the upgrade included putting WIFI and GPS sensors in the coaches so as to capture data about hard braking, sharp cornering, and excessive acceleration. A side effect was that we could display on a big screen precisely where all the buses were in real time (I know — this all sounds ridiculously low tech when we have Uber on our phones).
The day we switched on the system, we invited the CEO to check out the display, and he immediately zeroed in on a specific bus. “What is that bus doing there?”, he demanded, pointing at an icon of a bus down in the corner.
Turns out the driver had decided to visit a drive-through coffee shop, which was a strict no-no. Big bus coaches don’t interact well with overhead canopies, struggle to navigate through mall parking lots, and collide with lamp posts, other vehicles, and the innocent. Busted.
This was possibly my first modern experience with installing a sensor on a piece of equipment, and using the data from the sensor to take a decision (disciplinary action for the driver, unfortunately).
Data has been central to my work on digital innovations in oil and gas from the very beginning because digital innovation is almost always about capturing value from a reliable data asset. My simple digital innovation formula includes data at its very core:
- Sensors generate data
- Analytics (AI and machine learning tools) interpret the data
- Semi-autonomous and fully autonomous devices (robots, drones) carry out real work using that data
- Cloud computing is where data lives
- Trusted ledger services provide confidence in the data.
In addition to growth in their sheer numbers, sensors are evolving on many dimensions in parallel—shrinking in size and energy needs; collecting more data per unit time; collecting data more frequently; capturing different types of data (sound, visuals); working in increasingly varied places (deep underwater, high above the earth); carrying out analytics directly on the sensor platform; communicating that data in near real time; and, operating in a bidirectional mode (sending data collected to the cloud, while receiving software updates from the cloud).
As the cloud can now store virtually unlimited quantities of data, and carry out increasingly complex algorithms on that data, what was once scarce (data) is now in super abundance.
There is an ever widening gap between this exponentially growing data resource and the slowly adapting human-scale processes that actually generate value. The exceptionally rapid uptake of generative AI technology (which is based on data resources) highlights the cost of the gap. Closing this gap is at the heart of the creation of a center for data excellence.
Why Data Acumen Is So Low
In hindsight, it’s rather sensible that a center for data excellence is now part of the organization structure of a big oil and gas player. Without better data acumen, oil and gas companies are like street brawlers showing up for a fist fight only to discover the opponent is equipped with a pistol. We’re seriously outgunned in the fight for talent, the battle for capital investment, and the latest race called energy transition (which has a huge data theme to it).
The work culture of oil and gas is the basis for why our industry lags other sectors in how we think about and approach data.
The Reality of Data
- Data is not tangible, and oil and gas is the ultimate tangible industry. Professionals in the industry gravitate to the hard assets, be that physical plant, subsurface resources, or petroleum products.
- Data acumen is not part of the traditional curriculum for engineers studying the petroleum industry. The workforce is missing the modern skills for data management, let alone the basics.
- Data is often trapped in corporate and operational systems whose original designs did not anticipate data sharing and data access. Large oil and gas companies will have hundreds of distinct systems and siloed data, from hundreds of suppliers.
- Operational data is frequently in a time series, which poses its own special challenges (immense data sets with individual measures holding little perceived value, recorded without context, and held locally because of historic telecom constraints).
How People Feel About Data
- Oil and gas is a male dominated industry and the workers feel real deep pride in keeping its big physical assets running at pace. Workers gravitate to where they can capture those very human feelings. There is rarely such a similar feeling of pride and accomplishment from producing an accurate data set.
- The benefits of high quality, reliable data often accrue not to the originator of the data, but someone downstream of the originating data. For the originator, there is often no direct benefit to making data correct, so why bother?
- Individuals and teams resort to building and owning their own data and tools that they believe they can trust, which results in a proliferation of redundant data sets and tools, incurring a trust premium of slower reaction time and missed opportunity.
- Few if any industry participants have distinguished themselves or their business models based on the data they collect, so there are no role models in the industry for other firms to study and emulate.
How Organizations Treat Data
- There are no universally accepted standards for data definitions in oil and gas. Without standards, data is free to evolve in any and all directions without bounds. Once data proliferates freely, it’s hard to rein it in.
- Data is not treated as an asset by finance. Accounting principles lump data with plant and equipment. As such, it does not attract sufficient capital investment from the business, and control of capital is often how oil and gas professionals measure their self worth.
- The value of data is tied not to the value of the decisions based on that data, but on the storage device on which the data resides, or the value of the cloud services contract.
- The standard performance measures for oil and gas do not mention data quality directly. The metrics are safety (keeping employees, the assets, and the environment free from harm), cost (achieving low cost production since energy is a commodity), and reliability (producing the volumes of production committed in the financial plan). The link between being good with data and these performance measures is tenuous at best. Managers are not held accountable for data quality.
- Organizationally, oil and gas companies like to structure themselves along disciplinary lines, and create quasi competitions internally to surface the best capital opportunities. This structure stifles cooperation and data sharing, and leads to ‘roll your own’ data and tools.
Yep, it’s a mess. Frustratingly for digital transformationists, and fortunately for the industry, margins are often so good in oil and gas that these sins are simply viewed as an insignificant cost of doing business.
Maturing the Organization
I attended a Liquefied Natural Gas (LNG) conference last week, and a shipping company relayed an anecdote about converting one of their vessels from single fuel (marine bunker) to dual fuel. They did the conversion while the ship was in operation (versus being laid up in a dry dock for months).
Their lesson? Don’t try to swap out fuel systems while your engines are running.
Sound advice.
Unfortunately, we can’t just stand down a running petroleum business to fix problems. We need to do both.
Note that in the oil and gas data excellence context, there are four modes or general settings (ludicrous mode is not one of them).
- Gig mode — think of this as ‘every worker for themselves’, no reliance on others or central teams for any data or tools. Live and die by your own choices. I personally am a gig worker, and this is where I live. YouTube videos are my salvation.
- Sharing mode — the initial oil and gas response to the need for standardized behaviours. Some data is centralised into oceans, warehouses, or lakes, and some useful tools are shared across teams. A bit of tech transfer, some altruism, but no collaboration.
- Collaboration mode — a central technology authority promulgates a useful tool set, provides access to some learning and skill development, and enables cross team data sharing. Teams make data available on an as is basis, depending on resource availability and costs.
- Digital mode — a platform team identifies and supports a carefully selected set of data tools, a data academy program teaches data skills, and teams actively produce trusted data as a consumable for others.
Assess which mode you’re in, and which mode you need to be, and that will inform some of the moves you’ll need to make to mode up.
Switching on Digital Mode
Oil and gas companies have a long road ahead of them to get to digital mode, and not a lot of time. Here are some tactics that should be executed with haste.
- Create a data academy to educate the entirety of the organization on data.
- Set out your white lines or guiding principles about data, perhaps even as part of your organizational values.
- Establish and fund a data team whose role is to bring data thinking to the full organization.
- Achieve some quick wins by identifying and incorporating costly external data into a central data source.
- Build momentum by helping closely linked and mutually dependent teams find data to share.
- Put into place organization-wide data governance practices (performance measures) for ownership, search, quality and reliability.
- Create a data ecosystem where AI solutions will work reliably, such as choosing more open systems, embracing cloud computing, and adapting agile methods.
Conclusions
As I like to remind people, oil and gas products are hazardous, and we run this industry not on our physical exposure to the products but through the data we have about the products. That alone is a good enough reason to take data more seriously than we have.
Artwork courtesy of Geoffrey Cann
SHARE
Share This:
