IEEE Conference – Part 2: Business Card Exchange!

Facts written in this article could be slightly off, as these are my interpretations and understandings from my memory, which are not always accurate.

As promised in Part 1, I am writing this article within 24 hours.

Grady Booch

When we entered the conference hall, all time favorite author was talking on the subject: “Human/Ethical Aspects of Big Data”. He was saying that Big Data isn’t a new technology and in fact it’s one of the oldest technologies. It seems like in 1854 there was a big Cholera outbreak in London. Engineer John Snow mapped the cholera out broke areas. It indicated that cholera patients where concentrated in a one particular area of the city where there was a water pumping station. Once that station was shutdown, cholera got in to control.

Fig:  Big Data analysis of Cholera outbreak in 1854

He gave few interesting use cases for Big Data:

1. A small sensor can be installed on your car. This sensor can emit data on your driving pattern i.e. all the places you are travelling, the speed in which you are traveling….. Based on this data – your car insurance can be charged. i.e. if you go high speed, then insurance premium will go up and vice versa.
2. Similar sensor can be installed on the guns, to track your pattern on the gun usage.
3. During the maternity period, a sensor can be installed on mother’s stomach. This sensor can emit data each and every bit of data about the child’s behavior and patterns. Using this data, it possible to figure out whether baby is going to be born healthy or with any down syndromes. Its opens up option to the parents to take decision (like aborting the child …)
4. Since so much studies are going around genomic studies (with the help from Big Data), it could be possible to design the baby how you want to be born

Big Data is going to open up so much possibilities – we has as engineers has morale responsibility to let the bean counters (this is the word he used to refer to executives) – whether it’s the right thing to do for the future generations to come.

GE – Vice President

William Ruh, VP of General Electric Software was the next speaker. He made a point that when 1 billion people got connected, consumer internet was born. But currently in the world there are 50 billion machines. Their vision is to connect all these machines together and build what is called “Industrial Internet”.

Today GE is one of the largest aircraft manufacturers. So far they were getting only 3 data points from the aircrafts:

1. Takeoff Diagnostic Data (averaged)
2. Cruise Diagnostic Data (averaged)
3. Landing Diagnostic Data (averaged)

These data points, wasn’t helping them. Advent of Big Data, they started to get more than 200+ data points every few milliseconds from the aircrafts that where in motion. With this amount of data, they were able to identify a pattern. They were able to figure out that aircrafts which fly to marsh lands accumulate lot of dust in their fans. Once this dirt was cleared properly/frequently aircrafts flying to marsh lands started to have long life……

Fig: Data points reported from aircraft

He told that if they are able to reduce 1% inefficiency in their following lines of business, it would result in several billions of dollars savings (over 15 year period):

1. Frieght Rail – 27 Billion $
2. Health Care – 63 Billion $
3. Diagnostic Power – 66 Billion $

So with machines generating so much data, they are confident to identify the patterns and drive the efficiency. He told that in any business fuel is the most expensive cost. They are observing (or already observed) patterns that when rails to run at certain speed, it consumes less fuel. Thus by virtue of running rails above that threshold speed, they can bring this 1% reduction in the Freight Rail.

It seems like GE was in the 7th of 8th place in terms of Wind Turbine manufacturing. But now because of Big Data/analytics there are #1. They were able to lower the power generation cost from 40$ to a single digit number like 8$. Below is the slide which explains what they have done. Basically they have installed sensors with in each Wind turbine. Each Sensor is capable of communicating with other Wind Turbines in the neighborhood. Based on the wind blowing speed and direction, intelligence has been built in to these turbines, so that they can change their direction accordingly. Once any turbine in the neighborhood sets itself in optimum direction, it would communicate it’s direction to it’s peers. So other peer turbines would also adjust their direction accordingly. WOW, what a brilliance. There are also several other intelligence built in to these turbines as shown in the slide.

Fig: Wind Turbine Intelligence

Question that we should ask ourselves is: If to a mechanical instrument, so much intelligence can be built it, how much intelligence we can build to our software applications?

Note: After the presentation, GE VP walked down to his seat. Few minutes later when next speaker was talking, I saw Grady Booch walking to GE VP. He gave his business card and VP gave his business card in return. Booch didn’t do this with other speakers. I will leave it your speculation – what might happen after this business card exchange.

Kaiser Permanente – Director

Kaiser Permanente is a leading health care/hospital provider in California. It’s Director of Data Analysis – Michael S. Johnson was the next speaker. Here are few glimpses of his speech:

1. Size of Kaiser. Today they serve 8.7 million people with 17,000 physicians. They have 175,000 staffs, 37 hospitals and 611 medical offices.
2. At present they are dealing with 30 petabyte of data. Director acknowledged that this is not a big count in this conference, however he mentioned that their data is doubling every 2 years.
3. Director talked about few key initiatives. One of them was striking to me. When any patient is admitted in Intensive Care Unit for every few milliseconds several data points about his health are captured. If they start to exhibit pattern of any previous patient(s) who had serious illness- then their system would start to generate alerts. Based on that treatment can be given to the current patient, to prevent him from getting in to any serious illness.
4. Just by walking 30 minutes every day – one can lower his chances of heart attack by 50%, diabetes by 30% ….

I am going to take a break and write about the remaining (talks from Teradata, NetFlix, IBM, Google) in part-3. Best is yet to come. Stay tuned.