Thursday, August 24, 2017

John Roese CTO Dell Technologies Talks IIoT



I felt this video (8 minutes long) where John Roese (CTO for Dell Technologies) was interviewed at IoT World Solutions Congress really nailed it in terms of the important issues for IoT or more correctly, IIoT - Industrial Internet of Things.

Important points he brings out are:

  • Modernizing infrastructure will be key
  • Companies will need very flexible software to take advantage of IIoT
  • The industrial sector is ALL IN when it comes to IIoT
  • Collaboration is the key driving factor with IIoT


Wednesday, August 23, 2017

Oracle Might Let Java Go To An Open Source Foundation



Oracle's Aquarium blog by David Delabassee,  former long time Sun guy (Java Ambassador from early on and a thought leader at Sun) and now is the Software Evangelist at Oracle.   I am REALLY glad that David is still at Oracle and driving Java.   David is/was a great guy. David said:

"We continue to make great progress on Java EE 8. Specifications are nearly complete, and we expect to deliver the reference implementation this summer. As we approach the delivery of Java EE 8 and the JavaOne 2017 conference, we believe there is an opportunity to rethink how Java EE is developed in order to make it more agile and responsive to changing industry and technology demands.

Java EE is enormously successful, with a competitive market of compatible implementations, broad adoption of individual technologies, a huge ecosystem of frameworks and tools, and countless applications delivering value to enterprises and end users. But although Java EE is developed in open source with the participation of the Java EE community, often the process is not seen as being agile, flexible or open enough, particularly when compared to other open source communities. We’d like to do better.""

He also stressed the importance of Java's commitment to corporate customers:

"We intend to meet ongoing commitments to developers, end users, customers, technology consumers, technology contributors, partners and licensees. And we will support existing Java EE implementations and future implementations of Java EE 8. We will continue to participate in the future evolution of Java EE technologies. But we believe a more open process, that is not dependent on a single vendor as platform lead, will encourage greater participation and innovation, and will be in best interests of the community.   "

Looking forward to seeing how this rolls out!

Friday, August 18, 2017

Charlottesville by VICE and a Famous UVA Grad



These 22 minutes are a must watch if you have not see it.  Warning, NSFW - Not Suitable For Work because of language and violence.



Tina Fey said this better than anyone.


Thursday, August 17, 2017

Defense and Aerospace Plant Monitoring Using MTConnect Article by SME


I have worked a lot with the folks at SME - Society for Mechanical Engineers - and they do great work.  This article is especially well written with numerous real life examples.

Zooming ahead in data-management tool adoption by Sean Lyngaas - Contributing Editor 

 The article starts out:

"Manufacturers of many stripes can save money just by making better use of data emanating from the factory floor. But for those in aerospace and defense—Airbus Helicopters is a prime example—the potential payoffs are legion.

The combination of strict product-tracking requirements and high labor and maintenance costs are driving adoption of data management products
among aerospace and defense companies, industry insiders said."


 Aerospace and defense are seeing tremendous payoffs from knowing what is happening on the shop floor.  Dave McPhail nails it below:

“Every hour that I can save in taking it from a nonproductive hour to a productive hour is of substantially more benefit to aerospace and defense manufacturing than it is to, say, automotive or maybe food and beverage packaging,” said David McPhail, CEO of Ontario-based Memex Inc., which makes software that monitors machine efficiency. The aerospace industry involves expensive equipment, personnel, and product maintenance, which are all incentives to exploit shop-floor data, he noted."

Long time friend and MTConnect thought leader, RonPieper of TechSolve discusses MTConnect and a great IIoT example that TechSolve deployed for an engine aircraft maker:

"Another enticement is the fact that some aerospace companies are starting to require the MTConnect standard in their equipment purchase requirements, said Ron Pieper, product manager at TechSolve Inc., a Cincinnati-based manufacturing consultancy.

The ROI for aerospace companies adopting data-management tools is evident, Pieper and others said. He cited an example of an aircraft engine maker that wanted to monitor the consumption of a specific gas during the manufacturing process. After TechSolve installed sensors on the manufacturers’ gas lines, he said, they discovered a gas leak that had amounted to an annual loss of roughly $100,000."

The article addresses the multi-billion question of WHY aren't more companies embracing MTConnect and shop floor monitoring?

"Despite all of the incentives for greater use of shop-floor data products, aerospace and defense companies are not immune to the cultural hurdles practitioners say are impeding digitization in the manufacturing sector writ large. Some analysts estimate that the percentage of manufacturers that have implemented data-management software on the factory floor is still in or near the single digits."

My experience tells me that the number is in the VERY low single digits.

All of the individuals interviewed for this article nail the reason for slow adoption -- it is cultural and financial.  Below are the points made on culture and MOST importantly, Crawl, Walk and then Run in your implementations.

"Making the jump to data-driven manufacturing requires a culture within the company that sees data as the glue that can hold the factory together, McPhail said. The goal is “one objective view of exactly what’s going on in the factory,” he added.

One way of getting to that shared vision of data among management is to only bite off what you can chew.

“We actually caution shops, when I go talk to them about doing monitoring and data collection, about not trying to get too much data too soon, because the big hurdle is cultural; it’s not technical,” Pieper said.  McPhail echoed that reasoning, urging manufacturers to identify business objectives up front that data-driven manufacturing can help realize.

Of course, the flood of data available once factory machines are digitized can be overwhelming.
Jody Romanowski, CEO of software vendor Cimco Americas, said customers sometimes have grand ambitions for data collection—to want operators to scan, for example, dozens of downtime codes when machines aren’t running. Such high-volume scanning is often not feasible, so her firm works with customers to break down the amount of data sought into manageable categories.

“We’re always trying to find ways to make that happen more efficiently,” she said of crunching data captured on the factory floor. “That’s a huge consideration and still a struggle sometimes.”

To avoid flooding customers with data, Wintriss only dispatches data relevant to the customer’s mission, Finnerty said. “If we send a data word from one of our controllers up to the database, every bit in that word means something."

It is great to see MTConnect to continue to really grow and thrive.



Wednesday, August 16, 2017

Cisco's 75% of IoT Initiatives Fail Statement

There is an interesting article in ReadITQuick titled:

 
 In the article, Kulkarni brings out:
 
"Cisco recently conducted a survey to understand the Internet of Things (IoT) scenario, i.e., its hits and misses in an increasingly IoT-crazy technology world. The results were eye-openers to the reasons why over three-fourths of IoT projects were ending up as failures. A mere 26% of the undertaken projects were taken to successful completion, indicating that there is a great deal to be learnt and implemented in our IoT journeys. "

That number is really anything but surprising to me IF you factor out the manufacturing industry.  The reason I make this statement is that in manufacturing, IIoT or Industrial Internet of Things, is really about using sensors to augment what is already being monitored.  In other words, the framework is already in place and MTConnect is the protocol of choice for discrete manufacturing and IIoT is really the addition of sensors. 

Since IIoT is augmented to MTConnect, the ratio for success, IMHO, is in the 95% and up range.

 The survey was quite extensive as stated below:

"The survey collected the responses of about 1,845 IT bodies and the results were expressed in Cisco’s IoT World Forum in London, where Cisco chief executive Chuck Robbins talked about the problems that plague IoT forays by corporations. "

 The author brings out basic blocking and tackling projects of the failure - lack of commitment.  I don't care what project is, if you do not have a champion, forget folks!  Below they bring out the Holy Grail issue that I highlight.

"The first problem that he cited lay in the lack of buy-in in the IoT concept, leading to a lack of commitment to take projects to completion. In fact, a whopping 60% of the IoT projects are seen to stall at the proof of concept stage itself. The result is that enterprises are not willing to invest in the necessary IoT infrastructure, but merely want faster results by investing in a readymade integrated architecture that works “as is.” This is why most enterprises end up looking to buy IoT as a service, rather than taking the hard way and building a strong IoT architectural foundation. "

"Ready made integrated architecture that works "as is" " is what EVERYONE wants, but is really, really hard to do and why graduates in Computer Science quickly get to 6 figures of salary.  If it was easy, any damn fool could do it.  It ain't easy.  This is why I have always believed that MTConnect will be the tail that wags the IIoT dog for manufacturing.


 

Sunday, August 13, 2017

Joel Neidig of ITAMCO on the Industrial Internet of Things (IIoT)



I have often said that Joel Neidig of ITAMCO is the rock star of manufacturing.

Here is a very nice article on ITAMCO and what Joel is up to that was written by Mark Albert, Editorial Director for Modern Machine Shop, titled,

Connecting Forklifts to the Industrial Internet of Things

In this article, Mark brings out:

"ITAMCO, a manufacturer of precision-machined components and high-precision gears in Plymouth, Indiana, has a history of integrating its machinery and equipment with networked sensors and software. Many of these connections are powered by software applications for mobile devices—apps developed in-house by its own technology team. In 2012, the company implemented an MTConnect-enabled machine monitoring system. Soon after, key pieces of machinery were connected to the company's enterprise resource planning (ERP) system. Now ITAMCO has developed a communication system for its forklifts, citing this connection as a good example of how the Industrial Internet of Things (IIoT) will benefit manufacturing. In this case, it has made forklifts, the workhorses of the plant floor, more valuable than ever at ITAMCO."

Joel uses MTConnect extensively at ITAMCO.

What is absolutely worth watching is the 13 minute video at the end of the article where Joel discusses what ITAMCO does, including building the gears for a pump that was designed for New Orleans after Hurricane Katrina that can pump an entire olympic size swimming pool in 6 seconds!  That's right, 6 seconds. 

 

Wednesday, August 2, 2017

Exciting New Blog on Data Science -- Nina's Data Metrology World in Manufacturing

I consult for AMT - The Association For Manufacturing Technology and have the privilege of working with some extremely bright and passionate individuals at AMT.

AMT has always been the global thought leaders in manufacturing and this leadership position is accelerating.  It's as if there is a petri dish of collaboration between Silicon Valley and the world of manufacturing.

An example of this merge of technologies and ideas is in the very important area of data sciences and artificial intelligence (AI) branches such as machine learning.  To help make this critical transformation, AMT hired a very talented individual that is part of the MTInsight team.  MTInsight  is a leading platform for online business intelligence which delivers knowledge that leads to informed decision-making and increased productivity.  The individual's name who is leading edge data sciences effort is Nina and she is a Data Scientist.  Nina has spent the majority of her career within the quantitative realm.

Nina has started a very interesting blog titled, Nina's Data Metrology World in Manufacturing, which is a must read not only for those who have an interest in manufacturing and data, but anyone who wants to understand the right way to think about data.

Nina's has two very interesting blog posts:
I am looking forward to reading Nina's blog posts and learning!

How To Demonstrate Eight MTConnect Simulators To One MTConnect Agent


I was asked to update a set of MTConnect Hands-On Training Lab slides I did for a visit to Taiwan back in 2012.  Part of the update was making the MTConnect Simulator more interesting than just the single part being created that has been out at agent.mtconnect.org for a long time.  Don't get me wrong, it is a nice and simple simulator, with the binaries, data files and instruction out at github.com/MTConnect, but it was time for an update for those who want to dig into the details of MTConnect a little more and provide a little more sizzle as well.

This 12 minute video shows how to demonstrate eight MTConnect simulators to one MTConnect agent.

HUGE thanks to the great folks at NIST's Smart Manufacturing Systems (SMS) Testbed at National Institute of Standards and Technology (NIST) for making this data available at github.com.This is tremendously helpful for those of us in the MTConnect Community that NIST would do this.  Manufacturing is very fortunate to have thought leaders at NIST driving important ​test ​resources and data.  


I reference all of the info on how to run the simulator is out at github.com/MTConnect in the video.  You want to download MTConnect agent and simulator first, before you go through this modified version of the standard simulator demo.  You should take your time and go through the README after you download the above.  I also reference MEMEX's OPTime, which is free, and you can download it here.

Moving it to a Unix or Linux system would be trivial -- for you Unix folks.  Here is the link on my DropBox for the batch and config files to run the Eight MTConnect Simulators To One Agent.  Please note that I did modify the data to get the machine tools to start creating parts immediately.  What I specifically mean is that NIST started gathering data at 5am, but the machine tools were not making parts until 7am.  I simply removed the 2 hours of machine tools sitting idle for the purpose of this demo. 
 
Below is the 12 minute webinar I put together to show how to run this on your own Windows system.  



 

Any questions or comments, please use the comment section of my blog and I will be happy to help you out!


Tuesday, August 1, 2017

Pranab Chakraborty's blog post -- 3 Common Myths Around Machine Learning


This blog post Pranab Chakraborty's  blog post -- 3 Common Myths Around Machine Learning

The article leads off with a Bill Gates quote:

"A breakthrough in machine learning would be worth 10 Microsofts"

The article lays out the premise here:

"The resurrection of AI in recent years can be attributed to significant developments in machine learning systems, especially in one of its sub-field called – deep learning. Machine learning impartscomputers the ability to learn without being explicitly programmed”. Deep learning is a class of machine learning algorithms that use deep artificial neural networks with multiple hidden layers.
While evolution in machine learning drives the current AI boom, the hype has caused certain misconceptions around the capabilities of these systems. Some of these misconceptions have risen to the level of myths."

I won't spoil the punch line by listing all three here, but the author does make important distinctions between today's reinforcement learning and how a baby learns to walk.

"If we compare the learning process of a machine with that of a child, it becomes evident that machine learning is still in its infancy. For example, a baby doesn’t need to watch millions of other humans before it learns how to walk. She sets her own goal of walking, observes other humans around, intuitively creates her own learning strategy and refines that through trial and error until she succeeds. Without any outside intervention or guidance, a baby displays curiosity to learn and successfully walks, talks and understands others. Machines on the other hand requires guidance and support at each step of learning.

Moreover, a child easily combines inputs received through multiple sense organs to make the process of learning holistic and efficient. In one article, Dave Gershgorn indicates that “AI research has typically treated the ability to recognize images, identify noises, and understand text as three different problems, and built algorithms suited to each individual task.” Researchers from MIT and Google have published papers explaining the first steps on how a machine can be guided to synthesize and integrate inputs from multiple channels (sound, sight and text) to understand the world better."



I am excited about machine learning, but I am cautiously excited as I know at the end of the day it is still 1s and 0s running on hardware someplace and I remember the multiple AI winters going back to the AI Ambassadors in the 1980s at Sun Microsystems.