Predicting Production Delays and Expected Yields in Future Batches

Predikto will be releasing a new Advanced Analytics solution that will enable operators at continuous manufacturing facilities to predict production delays, asset failures, and expected production yield for future batches without the need to send us sensor data on a daily basis.

Operators will enter future batch settings like Steel Quality, Billets, Speed, Temperature, and a handful of other information.  Predikto will show in a simple graphical web form the risks embedded in the batch and the recommended settings to maximize yield and reduce the probability of having delays or failures with their equipment.  This could be a game changer. Stay tuned…


China Pollutes Less and Impacts Global Iron Ore Pricing

Iron Ore Pellet 2014

Some sources estimate that China has over 400 Steel Mills. Therefore, a shift to reduce pollution has lowered output from some “bad actor” mills in China reducing overall supply while the global demand for steel is expected to increase in 2014. As basic economics suggests, increase demand with lowered supply means higher prices. The Platts outlook report on global iron ore stated that pellet premiums are expected to remain firm into the first half of 2014 as supply growth struggles to keep pace with an increase in global demand.

Steel Mills are very susceptible to global demands and pricing fluctuations. Stay tuned for a new Advanced Analytics solution by Predikto to help steel mills predict asset delays while also predicting production yields by batch. Plant Managers and Maintenance Managers are constantly striving to maximize yield, reduce asset failures, and improve overall efficiencies. We believe Predikto is unique positioned to help Steel Mills make significant improvements to their bottom line.

Reactive Maintenance and the Telephone Booth

I was walking through the streets of New York recently and ran into a public telephone booth. I was amazed at my reaction since phone booths are almost extinct. Earlier that day my family went to the Museum of Natural History to see dinasours up close and personal, and it’s amazing how all these different and interconnected animals are extinct as well. It got me thinking as to what practices in industrial manufacturing would become extinct in the next decade or two.

The fact that 50% of the experienced manufacturing workforce will be retiring in the next 5 to 10 years is adding a lot of pressure to organizations to improve automation and depend less on these seasoned “experts”.

So if improvements in technology and automation are key to ensuring US manufacturing continues to stay ahead of the pack, then what areas will become extinct? I believe organizations focused on reactive maintenance will eventually become extinct. These reactive firms who have not implemented best practices around preventative or predictive maintenance will be gobbled up by their better performing competitors, will be acquired by PE firms who will then trim the excess fat, or simply go out of business.

What do you think? What areas of manufacturing or maintenance do you think will become extinct in the next decade or two?

Interesting Facts About the Railroad Industry

There are 7 Class I railroads in the US: BNSF Railway, CSX Transportation, Grand Trunk Corporation, Kansas City Southern Railway, Norfolk Southern Combined Railroad Subsidiaries, Soo Line Corporation, and Union Pacific Railroad.

Unlike trucks, barges, and airlines in the U.S., freight railroads do not strain the public purse. Privately owned railroads have spent $525 billion since 1980 building, maintaining and growing their 140,000-mile rail network. That amount equals 40 cents of every revenue dollar. Even during the economic downturn, America’s freight railroads spent approximately $20 billion annually to build and maintain the most efficient rail system in the world. In 2013, that investment is expected to increase to an estimated $24.5 billion, helping to keep America competitive.

Average inflation-adjusted rail rates (as measured by revenue per ton-mile) are down 44 percent through 2012. That means the average rail shipper can move nearly twice as much freight for the same price it paid 30 years ago — saving consumers billions of dollars in shipping costs each year.

Railroads are much safer. From 1980-2012, the train accident rate was reduced by 80 percent and the employee injury and illness rate fell by 85 percent. 2012 was the safest year ever for railroads, breaking the safety records set in 2011.

Railroads are stronger financially. Return on investment, which had been falling for decades, rose to 4.4 percent in the 1980s, 7.0 percent in the 1990s, and 8.5 percent from 2000 to 2011. That’s important, because railroad earnings today lead to rail investments in new locomotives, tracks, bridges, and more so taxpayers don’t have to.


Railroad sensors are a huge swimming pool of data

We like huge pools of data to tell interesting stories and help our clients improve safety and reliability. Railroad companies are capturing a lot of data and a potential client recently stated “We have a huge pool”.

Predikto met with a Class I railroad company in North America. They have 800 sensors installed across their thousands of miles of railroad track. The types of sensors included: – Hot wheel bearing detectors – Acoustic bearing detectors – Dragging Equipment Detectors – Wheel impact load detectors (WILD) – Cracked Wheel Detection (Machine Visioning)

The entire railroad industry has heavily invested in sensors and technology to improve reliability, safety, and velocity. 2012 was the highest safety track record in railroad history. Clearly, these guys are doing something right. What is amazing is the wealth of data they are capturing and the lack of predictive analytics to enable additional reliability improvements. We should not be surprised since accurate, relevant, and actionable predictive analytics is very challenging. Predikto is finding a huge void in organizations of all shapes and sizes in their ability to tell a story (more specifically, to predict future reliability) by carefully dissecting and analyzing their own data. We see this as a great opportunity to make an impact and improve railroad safety, reliability, and improve operating efficiency.


Modeling & Simulation to Keep US Manufacturing Competitive

Today I read a 2009 white paper on where it explains why US Manufacturing needs to focus on technological breakthroughs using High Performance Computing (HPC), modeling, and simulation to find better and smarter ways to improve manufacturing.  The article describes how transformative technology is used by international competitors to US manufacturers. It provides an example of a HPC German lab helped the nation’s coal fired plants use modeling and simulation to to optimize plant design and operations.

I fully agree with many points in the article.  Manufacturing companies are being asked to perform more with less resources and budgets.  Many manufacturing companies are experiencing higher rates of failure due to degrading and old equipment.  The use of technology like modeling, simulations, and predictive analytics can have a significant impact on operations and asset reliability.


Non Productive Time costs drilling contractors $100 to $150 million a year

We recently saw a study that Oil & Gas drilling contractors spend $100 to $150 million dollars a year in non productive time caused by failure in their equipment.

Predictive analytics is an area of statistical analysis that deals with extracting information from data and using it to predict future trends and behavior patterns. The core of predictive analytics relies on capturing relationships between explanatory variables and the predictive variables from past occurrences, and exploiting it to predict future outcomes.

Predictive modeling draws from statistics and optimization techniques to extract accurate information from large volumes of data. Modeling techniques produce interpretable information allowing maintenance personnel to understand the implications of events, enabling them to take action based on these implications.

In the maintenance industry, predictive analytics builds on prior investments in enterprise asset management (EAM) systems, combines real-time data from sensors and other acquisition techniques with historical data to predict potential asset failures, and enables the move from reactive (scheduled, break-fix) to proactive (condition-based, preventive) maintenance.

Predikto is able to create models to predict failures in similar equipment.