You are Home   »   News   »   View Article

Becoming data driven and agile

Thursday, August 13, 2015

Industries around the world are becoming increasingly digital - data driven and agile - adapting to the new style of business. Martin Houghton, Chief Data Officer with Hewlett Packard explained how this transformation could be applied in the oil and gas industry.

"You've heard the term, 'data is the new oil,'' said Martin Houghton, chief data officer with Hewlett Packard, speaking at the Finding Petroleum forum in Aberdeen on March 17.

"We have a term 'big data changes everything'. It really is changing everything in many industries."

Mr Houghton is a former leader in global data standards and business intelligence in BP and Information Management and ERP manager with Shell.

"I've noticed oil and gas are relatively slow to come to this," he said. "But oil and gas is an industry that creates a huge amount of highly varied data. Digitisation of the oil field has created lots of data, but in order to become truly data driven, energy companies need to understand how they can leverage 100% of the data. Everything from video, voice, text, documents, seismic, essentially every form of digital content needs to be utilised to create valuable insights, and this data needs to be made available to 100% of the workforce.'

In the future, the most useful data will form "systems of insight", where data is combined in new ways to tell you something you didn't know before, he said. 'It's game changing rather than transactional, triangulating digital content - bringing together data in new ways to find out something new' he said.

Oil and gas companies can use techniques like this to work out how to get more out of seismic data, leverage that with other forms of data such as fibre optic, FTG, electro-magnetic, video etc. and work out how to optimise their assets, understand HSEE incidents, proactive maintenance and maximise recovery.

Mr Houghton suggests that oil and gas companies should ask technology providers like HP to look after their technology needs, allowing them to focus on their core business. With expertise spanning many industries and this HP's experience can help you oil and gas companies to innovate by applying techniques from other industries to create new opportunities. Essentially reusing examples, that when applied to a new scenario, become innovation.

Access to data

A first step to creating 'systems of insight' is getting 100% your data together. "People are talking about the data lake, a big blob of information,' he said.

But not many companies have a system to bring all their data together.

"How many people know where all the documents they created during their career are?" he asked. "That's knowledge and information not shared, essentially lost.'

For example, 'Bob is in the Gulf of Mexico and a paper he wrote last year contains vital information that could help you with a current challenge, problem is, you don't even know Bob exists, let alone that he wrote the paper"

Many new industry entrants, often new graduates, are surprised to learn that they can't find company information they need. 'People's expectations coming into industry are changing rapidly. They expect information to just be there," he said.

Even the regulators are demanding information to be instantly available. "I ran an area of regulatory reporting for an oil major. The auditor would contact me two months in advance. Now the regulators can knock on the door, ask you anything and expect an answer in 24 hours'

Many companies have been trying to integrate their operations by integrating their software applications and processes, he said. But perhaps just trying to integrate data is better. "There are hundreds, potentially thousands of applications. You can't possibly integrate all the systems, but you can integrate the data," he said.

A data lake can help replace traditional ways of working where there is generally point to point collaboration between specific departments but no overall company data store.

For example rigs would share data with their local office but not with company headquarters, and people saving their work on their personal computers until they need to share it. So basically people still working the way they did before computers came along with personal files and content not made available for the benefit of the whole enterprise.

'The old model of the world is one where there is data everywhere, but with very little insight,' he said. "That's overstating it, I realise, but I hear this all the time.'

One forum delegate stated that people sometimes prefer to use their own data because they don't trust work someone else has done.

Mr Houghton said that these systems aren't about changing the way people work, but providing access to the company's best most relevant information and data.

Data lab

Once you have your data lake, the next step is to try to do something with it. 'We can start to build analytical applications that can harness 100% of the data."

One way to do it is to build a company data lab. 'It's an experimental discovery environment with the purpose of finding out new things," he said. "Very different to traditional analytics type models."

"You've seen maps in films where they try to locate the criminal by triangulating different dots. That's what these labs do, bringing together all kinds of digital content - audio and video can be as relevant for doing analytics as structured data - to show you things you haven't seen before; correlations and patterns that can reveal new opportunities."

HP builds tools where you can 'immerse yourself in the data and insights, discover and find things that are actually happening, create useful reports and visualisations,' he said.

The data lab doesn't have to be a physical place, because people can access the same system anywhere in the world. Often companies find that its two best experts in a certain sort of geology are halfway around the world from each other.

Usually you need both domain professionals, such as geologists, and data scientists, working together. But some individuals count as both.

'There are geologists coming out of university who already know how to write in the statistical computing language 'R', he said. 'People are graduating with different skill sets than we've had in the past."

Oil and gas companies often dismiss things which did not work the first time but data tools also give you provide the opportunity to try out things many times in different ways.

"Often these things are deployed for a specific business case, and that business case has a particular output, and if your exploration doesn't produce that output it can be deemed a failure."

With data analysis techniques, you do some analysis and see if it tells you something interesting or not. and If it doesn't tell you something interesting, you try something else.

"We need to have an open mind to failure, it's not a negative thing," he said.

Old pdfs

One audience member said that 90 per cent of the company's data is in unreadable, very old pdf files.

Mr Houghton replied that HP has technology which can automatically scan and index the pdf files and see what useful information can be taken from them.

"We use the term 'dark data', he said. 'We get lots of oil and gas companies saying, we've got a big server array with petabytes of content, we know we need to do something about it but we don't know what it is, a giant blob of stuff." The challenge here is to reveal what the content is so the right action can be taken. Value it, delete it, and leverage it for further insight and discovery.'

Examples of analytics

HP was involved in one analytics project for an oil major to try to understand the social and political landscape impacting on oil prices, and try to make oil price predictions, Mr Houghton said.

Working on a project for a global agri-chemical business HP analysed data from drones used to monitor crops

Martin, I don't get the 'so-what?' for the O&G industry with the Nascar example? The relevance needs to be shown, or example removed

HP built an analytics system was for US racing company NASCAR, to help the company analyse all the comments made on social media while races are going on, so company staff can get a better understanding of what the fans are thinking about.

NASCAR has a "Fan Engagement Centre" where all the data is gathered together and analysed. The solution includes hardware (powerful servers to store and process data) and software, such as HAVEN, to extract meaning from the data. The data is then displayed to staff on video walls.

It receives data from Twitter, other social media sites, and media outlets. The company can track what the fans are thinking about as major announcements are being made, such as a team producing a new car.

The aim is to provide 'actionable data' - something which can actually be used, Mr Houghton said.

This is relevant to oil and gas, because companies also want to get insights from a wide range of different data.

Associated Companies
» HP


To attend our free events, receive our newsletter, and receive the free colour Digital Energy Journal.


Latest Edition Mar-Apr 2024
Apr 2024

Download latest and back issues


Learn more about supporting Digital Energy Journal