How can machine intelligence create value for business?

As part of Deloitte’s eighth annual Technology Trends report, we sat down with Mark Sheppard, CIO at GE Digital to get his take on how Machine Intelligence can create value for business. Here’s what he had to say.

The industry is seeing a resurgence of machine intelligence enabled by today’s substantially increased computer power speed, supported by a new set of applications and an increased general willingness and openness to change from our clients. Forty years ago, learning from computers was ‘scary’, but today programs such as Siri and Alexa are household names.

In November 2016 GE acquired machine learning company, recognising the tremendous potential value of machine intelligence, deep learning and cognitive analytics in the ‘industrial internet’ to drive productivity and optimisation, such as in the Australian mining and oil and gas landscape. The concept of using sensor technology, big-data analytics and remote cloud networks to drive performance improvement is becoming increasingly important to the energy sector in recent years. It is arguably the confluence of these three things that make efficiencies possible: individually they are important, but it is bringing them to together that can make a real difference.

A little goes a long way: 1% improvement = $1.8 billion (AU)

An example of the impact that can be made is around the issue of the derailing of locomotives pulling freight like iron ore or coal. It can take days to clear a derailing, which has a significant impact on production, productivity and the bottom line. Based on information received from sensors on a locomotive (on bearings, or to measure fuel or weight) you can effectively sense if it is going to break down and take action to prevent this.

Some 900 sensors can produce one terabyte of data per hour. Some analysis of the data occurs immediately on the locomotive, some in the cloud, using edge and cloud based computing and the remaining data is uploaded when it pulls into the yard. It is the power of machine learning that can make a 1% improvement on the bottom line. This might not sound like a lot until you consider that a 1% reduction in system inefficiency can translate to savings of around $1.8 billion (AU) annually.

The amount of data and insight radically changes the value proposition from IT: we are now selling an outcome which is the reliable movement of iron ore, rather than the packages of software and hardware. This changes our value proposition in the eyes of our clients.

Making a social impact

Machine intelligence is creating a fundamental change to people’s lives at the grass roots level. It is being applied in the agricultural sector to solve real human problems and the social flow-on effects have been remarkable.

For example, GE is working with an Indonesian company to help farmers improve yield on their small holdings. These farmers often struggle to stay above the poverty line, particularly when on land where size and boundaries are not clearly defined. Further, on average, 0.5ha is what they have to work with, yet they need 0.7ha to stay above the poverty line.

A cloud based app product called Hara, sitting on top of a GE platform called Predix, has been developed to help the landowner tag the corners of the land, as well as feed regular soil sample data into the app every two weeks. This data is coupled with satellite technology which reviews the leaf shape, size and colour for that plot, and provides advice on what fertilisers and pesticides to use, and when they should be used.

This has resulted in around an 80% increase in yield, which is having a significant impact on the wellbeing of the farmer. Two further developments have been that the banks will now enter into micro finance agreements with farmers allowing them to expand, and the machine app is able to bundle the fertiliser requirements with other farmers.

This has encouraged the bulk purchases of required fertiliser as a co-operative, which has further driven the costs down.

Advice to CIOs

In my experience, a lot of CIOs may have been burnt with big projects, and so there is a reluctance to try new things.

My advice is that machine intelligence is one area where CIOs need to think like a start-up: experiment, launch and learn. The amount of savings that can be achieved out there far eclipse the cost of these technologies. Give it a go, and treat it as a digital transformation and part of the drive for organisational productivity.


Mark Sheppard is the Vice President and Chief Commercial Officer for the Asia Pacific region of GE Digital.

GE Digital drives the development and delivery of GE’s software strategy across the organisation.

Prior to his current role, Mark was the Chief Information Officer (CIO) for GE in Australia and New Zealand, as well as the CIO for GE Global Mining. Mark’s vision for cultivating how technology supports the industry has been instrumental to the growth of GE’s Energy, Oil and Gas, Aviation, Healthcare and Financial Services businesses, in one of the company’s largest and fastest growing regions.

Before joining GE, Mark lectured in Computer Science in the UK, after having completed his PhD studies in Artificial Intelligence.

Mark specialized in the use of neural networks in the analysis of big industrial data and computational linguistics.


Want to stay up-to-date?

Stay on trend and in the know when you sign up for our latest content