GE Joins the Ranks Of Companies Relying on Big Data Analytics

Bill Ruh, GE

GE, a forerunner in the UK manufacturer’s market, is utilizing big data analytics to stay ahead of the curve. The analytics is allowing GE to predict the need for maintenance before it can disrupt the services.

Apart from the electronics that the company is known for, it notably manufactures turbines, jet engines and medical scanners. Running such large scale machines requires numerous sensors that collect data on engines and machines. This data can then be analyzed for a pattern.

When large machines come into play, small failures in parts or disruptions can lead to major downtime to get the machines back up and running. GE is relying on big data analytics to provide services as a part and parcel of the product. The services increases efficiency by allowing the machine to adapt to changing environments continually.

Bill Ruh – The Man behind Innovations

Bill Ruh, the Vice-President for software at GE research has been credited with thinking of innovative ways to put the customer first. The business opportunities that GE has strategized have surfaced by looking through the perspective of the customer.

For example, Bill Ruh said that GE enables the airline industry to save on fuel costs through their software. The technological advancement allows airline pilots to monitor and manage fuel efficiency in real time. The motivation behind creating the technology arose from the fact that the airlines spend nearly 200 billion dollars on fuel every year. A savings of only 2 percent was a substantial 2 billion dollars.

The movement planner technology also by GE allows train drivers to use the cruise control while being fuel efficient. The software utilizes the data of the location of the train, the terrain and the data is converted into the optimal speed at which the train can move while retaining fuel efficiency.

Vegetation causes more power lines to snap than any other reason. This thought provoked GE to create a software that allows pruning of trees along major power lines in a cost effective manner. The software is currently being utilized by a Canadian electricity supplier.

While most companies focus on bringing newer technologies, GE is focusing on increasing the efficiency and servicing equipment. The heavy machinery industry is seeing large benefits from the GE technology with nearly zero downtime and large savings.

Bill Ruh, snagged from Cisco nearly two years ago was quoted as saying, “We invested $1.5bn over four years to develop services and create new software. We are working on making devices more intelligent using sensors; and controllers that can be configured in real time.”

The use of the technology is best explained through a wind farm. The turbines that are in front tend to affect those behind them. Vibrations cause stress fractures on the turbine blades which then shut down and require downtime for repairs. GE’s software can allow experts to adjust the turbines based on their positions and the wind. These adjustments can help avoid the vibrations that are the root cause of the stress fractures on the turbine blades. Through the use of the software, GE allows wind farms to see a 2 to 5 % increase in the wind farm efficiency.

For an idea of the sheer amount of data that the software processes, one need only compare the data that the turbine sensors generate in contrast to the twitter feed data over the span of a single day. An entire day’s twitter real time feed amounts to 80 GB. Each blade on a gas turbine engine has a sensor that generates nearly 520 GB every single day and there are an average of 20 such sensors. Big data analytics is therefore no mean feat.

Datacenter automation was never the focus in the industrial world. Self-healing and autonomous computing is the aim of the IT systems management tools. While these tools were denied access to the heavy machinery world, they are beginning to find a niche in the arena. The industrial world is changing by the minute. People with years of experience are retiring. The inexperienced but educated replacements do not have the same intuition or experience. The demand for predictive analytics technology is therefore on the rise.

Understanding the Technology

Although the sensors are connected via the internet, the functioning of that internet is vastly different from the functioning of the internet that allows normal web browsing. Bill Ruh explained when he said, “The internet is optimized for transactions, but in machine-to-machine communications there is a greater need for real time and much larger datasets. We are seeing more processing on machines due to latency.” He was of the opinion that the current state of cloud computing as it is, is not suitable for machine to machine or M2M interactions.

Built-in memory database systems form the core of the technology used by GE. As mentioned earlier the size of these datasets are significantly large and GE is relying on NoSQL and Hadoop currently. Apart from these analytics software, GE also has its own software for the time series analysis. Despite having a range of software, GE is also working in tandem with Amazon web services and Microsoft azure to bring cloud computing into the picture.

Humanoid Machines

The machine is becoming more and more human like as technology is progressing. While the sensors can locate their own positions, they can also, very much like humans upload their status updates. The machines can also become ‘friends’ on a social network similar to humans.

For example, an engine that has GE sensors on an aircraft can alert the ground staff of the aircraft landing via social network of engineer friends. The engineers armed with the knowledge that the sensors have passed on, can carry out any repairs or maintenance as suggested by the data. This network also increases the efficiency of big data analytics.

 Security Threats

As machines get more and more intelligent, so do humans in their manipulation of the technology. Intelligence in machines and M2M communications can be easily exploited by a StuxNet attack. The worst kind of threat according to Ruh, will arise from the inside.

The machines will need an authentication and role-based security that is sophisticated and capable to holding off threats to the system. The human accessing the machine’s controls needs to be understood by the machine. The kind of security associated with commercial operating systems is the requirement in M2M interactions as well.

Ruh is of the opinion that the management of the iOS security is the way to go. He is looking to implement a sandbox approach to security.

User Interface

GE relies on a standardized HTML5 on the user interface side of the software. The internal working of the software however still uses Linux, Windows and Android operating systems. In order to make the software more user friendly, Ruh predicts that apps much like the ones found on android and iPhone systems will be in use shortly on the machines. The apps according to Ruh will extend the functionality of the entire machine.

If apps were to be used, the programming model will be vastly different from what is in use currently. The security framework that is also currently in use will need to be strengthened and made more robust before the launch of apps on the machines.

The Future

Predicting the future is a task that engineers in GE are constantly engaged in. A 200- person site in the Silicon Valley is specialized in developing services for GE. The services that are being designed will increase efficiency, as they are powered by sensor networks and software. The technology provides real time controls and carries forward a vision of zero downtime.

The Silicon Valley center is primarily a software center but it is also a programming unit that is working to develop the new services. The company is seeking new skills. The new hires have to be data scientists, a skill base that is in demand across the world. Nearly 1.5 million data scientists are in demand for positions around the world.

It is not only mathematicians or engineers that are in demand, but it is analysts of sorts that the company is looking for. People who can understand data sets in relativity to the machines from which the data originates in the real world.

Bill Ruh, was quoted as saying, “The technology providers gave us a solution to business intelligence but analytics is a lot harder than people think. Ten to 15 years ago people were told BI gives greater insight, but the BI tools have been used just for better reporting.” The logic for hiring analysts according to Ruh stems from the rejection of the BI tools.

In the game between machines and humans, IBM made sure that the machine won. A game show called Jeopardy in the US was played between a super computer called Watson and a human being. The computer won hands down. Such super computers however is not what IBM is looking for. Despite being sophisticated and programmed to handle any task, Bill Ruh is of the opinion that the computer is not on par with the analysts who can answer the real world questions better. Ruh was of the opinion that the super computer Watson cannot predict when a machine will break down.

Along with a new and motivated set of employees, GE is looking to leave a large footprint in the world of big data analytics at Cambridge, UK.

Latest Columns

Rural BPO’s and Vision of Mahatma Gandhi

The main advantages enjoyed by the owners of rural BPOs are lower investment in infrastructure, lower salary and they are not affected by huge attrition as in the case of urban BPOs. They are not worried over other overhead cost such as transportation and cafeterias. Since back office works does not create huge revenue, it […]

How IT team should become a business growth enabler?

The IT team of any corporation can be considered playing a vital role as a business growth enabler with respect to the revenues. However, but the IT team has been failing to promote business growth. A new survey by Economist Intelligence Unit and Juniper Networks suggest it is the strategic transformation in departments that the […]

Speak Your Mind