by Gary Mintchell | Dec 15, 2016 | News, Operations Management, Software
Data Science has gotten us to the point of collecting servers full of manufacturing data. We can do some analytics. But there are miles to go before we sleep.
This press release crossed my email stream last week. I haven’t time to interview the founder–that will come later. But here is a teaser.
Data Science Pioneer Drew Conway Closes $2.5M in Seed Funding to Bring Machine Learning to Industrial Operations
New venture Alluvium delivers “Mesh Intelligence” to close the machine-to-human gap
Alluvium, developers of Mesh Intelligence solutions that harness machine learning insights for real-time applications in industrial use cases, today announced $2.5 million in seed funding led by investors IA Ventures, Lux Capital, and Bloomberg Beta. The machine learning venture is running pilot projects of its Mesh Intelligence technology in fleet management, and oil and gas, among other vertical industrial applications.
Alluvium aims to conquer one of big data’s greatest unsolved challenges for complex industrial operations with expert human operators. Alluvium’s breakthrough Mesh Intelligence solution frees the data from these proprietary systems, transforms it into rich information streams, and provides real-time insights to human operators for immediate action.
“The commoditized big data stack is fundamentally broken for complex industrial operations,” said Drew Conway, Founder and CEO at Alluvium. “Modern industrial assets and hardware are continuing to be instrumented by OEMs who have not considered how these heterogeneous streams of machine data should be leveraged in the overall workflow and data strategy of the organization. And the modern analytics ‘stack’ — where data is moved and crunched in back end systems — does not meet the real-time requirements of human operators at the edge.”
Conway, who earned his PhD at NYU, is a leading expert in the application of computational methods to social and behavioral problems at large-scale. He started his career in counter-terrorism as a computational social scientist in the U.S. intelligence community and is known for his venn diagram definition of data science as well as applying data science to study human decision making.
At the core of Alluvium’s Mesh Intelligence platform is unique technology for extracting data from all elements of complex industrial operations — tablets, sensors, as well as industry-specific assets — with no expectations of compute resources or network bandwidth. This breakthrough allows machine learning processing to occur at the edge of systems where human operators need data most — in-real time.
“The early days of big data were about capturing and storing the vast amounts of new information streaming from devices in manufacturing, transport, medicine and more,” said Mike Olson, co-founder and Chief Strategy Officer at Cloudera, and a seed investor in Alluvium. “As that technology has matured, the more important and more interesting problem has become: What can we learn from all that data? Alluvium is focused on extracting meaning from streaming data coming from hardware that instruments all sorts of industries. The company augments human expertise with its powerful machine learning technology to make customers smarter and help them operate better.”
Independent research and surveys show the massive economic opportunity for IoT and machine learning across industrial use cases. A report by Jabil found that “$1.9 trillion dollars of economic value could be created by the use of IoT devices and asset tracking solutions.” For U.S. oil and gas suppliers — an industry where Alluvium has had significant early traction — the daily cost of unplanned downtime at a refinery can reach $1.7 million per day, and the daily cost of unplanned downtime for liquid natural gas drillers can top $11 million per day. A recent McKinsey report found that “car data monetization could be as high as $750 billion by 2030” — which has far-reaching implications for fleet management. Analyst firm Gartner forecasted more than 6 billion connected devices will be in use worldwide in 2016 supporting more than $265 billion in services. And in a 2015 “Moving Toward the Future of the Industrial Internet” report by GE and Accenture, 84% of executives expected Big Data to shift the competitive landscape within the next year.
“Bringing machine intelligence into the physical world is an incredibly difficult task,” said Shivon Zilis, partner at Bloomberg Beta. “We were excited to back Alluvium because of their unique insights into how complex industrial systems could be transformed by predictive engines.”
Learn More
Read Alluvium Founder’s Perspective on Starting the Company
by Gary Mintchell | Jun 23, 2016 | Automation, Internet of Things, News, Process Control
This week Honeywell Process Solutions held its 41st annual Honeywell Users Group (HUG) for the Americas in San Antonio. Speakers were all over the Industrial Internet of Things (IIoT) trend.
I did not make it. There are too many places to go, and I no longer work for a larger company with a financial base for lots of travel. So, I stayed in Sidney. News came via HPS’s PR agency and Control Global. Walt Boyes posted some cryptic notes on his blog one or two days.
Vimal Kapur, president of HPS, told attendees, “The Industrial Internet of Things will allow manufacturers to more efficiently gather and analyze a broader range of data across multiple operations and plants to use data to transform entire enterprises.”
Showing how IIoT is working in real-world scenarios is especially critical this year. A recent survey of North American manufacturing executives suggests that investments in data analytics are rising. However, companies are still grappling to better understand its benefits.
The Honeywell-commissioned survey, conducted by KRC Research, found the majority of respondents (between 60 and 70 percent) believe data analytics can help reduce equipment breakdowns and unplanned downtime. It can also help reduce supply chain management issues. This is noteworthy because around 40 percent of the executives see unscheduled downtime and supply chain management issues as the top two threats to maximizing revenue.
At the same time, though, nearly half of the respondents said they don’t fully understand the benefits of data analytics. More than a quarter say more proof is needed to show that data analytics work.
IIoT to Analytics
I find it fascinating how quickly the news from HUG transitioned from IIoT to data analytics. Some people look at the IIoT phenomenon as connected edge devices often through a network using the IP protocol. Increasingly analysts are broadening the scope to include the entire system of connected sensors, data collection, analytics, visualization.
That automation companies, such as HPS, are now emphasizing data science over control and instrumentation is an intriguing proposition to ponder over the future.
Digital Transformation
Just as I witnessed in Hannover, the real technology and term people are concerned with is digitalization.
“We are seeing a lot of interest from our customers attending this conference on how best to manage this digital transformation within their companies. They are looking to get more production out of existing assets and determine the best areas for investment for the long-term success of their operations – to improve process safety, reliability, security and sustainability,” said Kapur.
“HPS has been providing solutions and technologies to help manufacturers leverage critical operational data on a site by site basis for years,” said Andrew Hird, vice president and general manager of HPS’ Digital Transformation business. “Today, with the increased connectivity and the IIoT by Honeywell, they are able to gather and analyze data across multiple sites to find and implement best practices. The results of the IIoT survey of manufacturers reflect very closely the areas where Honeywell has been investing in solutions.”
Product Introductions
Honeywell’s newest industrial automation technologies unveiled at HUG included:
- Data acquisition and analytics: The expanded Uniformance Suite software provides real-time digital intelligence through advanced process and event data collection, asset-centric analytics and powerful visualization technology, turning plant data into actionable information to enable smart operations. This solution is a backbone for the IIoT by Honeywell. New this year is Uniformance Insight, which allows customers to visualize process conditions and investigate events from any web browser.
- Control: ControlEdge PLC, one of Honeywell’s first IIoT-ready controller, is part of Honeywell’s next generation of controllers providing unprecedented connectivity through all levels of process and business operations. When combined with Experion, ControlEdge PLC provides secure connectivity and tight integration to devices from multiple vendors and works with any SCADA system. Also showcased is the newest version of Experion Orion that will help industrial plants further optimize automation project execution, reduce loop commissioning time, minimize operational risk and protect intellectual investments while keeping current with today’s technology.
- Natural gas measurement: Honeywell’s new IIoT-ready gas measurement and data management solutions for North American gas transmission and distribution from the wellhead to the burner tip. The portfolio includes the EC 350 PTZ Gas Volume Corrector, Honeywell Elster Rotary Gas Meter, the Cloud Link 4G Modem, MasterLink and PowerSpring.
- Connectivity: MatrikonOPC Unified Architecture (OPC UA) extends the highly successful OPC communication protocol, enabling data acquisition and information modeling and communication between the plant floor and the enterprise reliably and securely, accelerating the IIoT.
- Mobility: Honeywell Pulse is a new mobile app used to remotely connect plant managers, supervisors and engineering staff to customized real-time plant performance notifications sent from HPS’ industrial automation software. It brings relevant metrics and the tools to resolve issues directly to their mobile device.
by Gary Mintchell | Jun 9, 2016 | News, Productivity, Workforce
Industry and manufacturing leaders recognize the trend to the next step in the evolution of enterprise effectiveness and success. The industrial digital revolution is an overnight sensation that has been 30 years in the making. We began with digital controls then adding human interface and then information handling.
Internet of Things with its proliferation of sensors and other smart edge devices, IP networking, data science, and advanced analytics (business intelligence) combined take us to a whole new level of enterprise effectiveness.
The trite question from marketing people often goes, “What’s keeping our customers awake at night?”
Well, are executive worried about the capability of technology?
Two research reports just came my direction recently from a couple of my go-to sources for what’s happening with the thinking in the industrial/manufacturing executive suite. One is from PwC, What’s Next in Manufacturing: Building an Industrial Digital Ecosystem, and the other from Accenture Digital Skills Gap Slows Manufacturers’ Push to Build Digital Factories.
No, it’s not technology that worries them. First it’s people and culture. Are there sufficient people with digital skills? Will the culture make the transition? Then, of course, they worry about how large the investment might become and what the return will be. It’s people and economics.
PwC Digital Industrial Survey
In this report, PwC shares results from a survey of global industrial products companies, shedding light on what manufacturers are doing now to build out their digital operations and what bottom-line benefits they expect to yield through those efforts.
Buying into digital: manufacturers plan to ramp up investments
In the last two years, US manufacturers invested an average 2.6% of their annual revenue in digital technologies. In the next five years, they expect to lift that investment to 4.7% of revenue—for an estimated $350 billion in investments in digital operation technologies across automotive, industrial production and manufacturing industries alone.
Venture capital funds flowing, too
Since 2011, some $3.6 billion has poured into VC-backed start-ups across a selection of digital technology sub-sectors, with investment rising at a 47% clip–more than double the annual growth of total VC funding (18%) in all sectors over the same period.
“Digital deals” have comprised 15% of all US M&A activity since 2012
According to a PwC/Strategy& analysis, more than $6.0 billion has been invested on “digital deals” in North America alone since 2012, comprising some 15% of all M&A deals over that period.
The greatest challenge to a “digital vision” is cultural
In the context of embracing digital operations technology, three of the top 10 challenge areas identified by surveyed companies relate to organizational readiness and financial concerns. Some companies anticipate high investment requirements with unclear return on investment, and lack of digital standards and issues related to data security and intellectual property are also noted.
Monetizing digital operations sought through cost reductions, revenue generation
Nearly two-thirds of manufacturers expect that adopting digital manufacturing technologies will translate into lowering operating costs by at least 11% mostly via efficiencies through automating processes and production. Meanwhile, over half of these manufacturers expect such adoption to boost revenues by at least 11%.
How digital technologies drive bottom-line results Manufacturers are just scratching the surface of monetizing digital manufacturing. Some key drivers to achieving cost-cutting and revenue uplift from digitization with the introduction of smart, connected manufacturing technologies and products and services include:
- Lowered “price of variability” across production and processes
- Moving from analogue products to “connected, digital products”
- Manufacturing data…and new business models
- Software-enabled upgrades to products
- Pay-as-you-go model
Building a digital manufacturing strategy
Building a digital strategy requires a thorough self-assessment to determine a company’s “current state” of its digital evolution—and, just important, defining its “target state”. This means tailoring digital operations solutions to a business’ assets and making the right moves at the right time—from ramping up data analytics capabilities, to monetizing product data to considering a “digital deal”.
PwC concludes, “The future of digital manufacturing holds many “what-ifs”. But, if it unfolds as dramatically as our survey indicates, most all manufacturers will be altered to some degree. And, for every “what if”, there are choices manufacturers ought to consider.”
Accenture Researches Industrial Digital
Take a look at some of the results of Accenture’s research. Although the majority ofmanufacturers have implemented digital platforms, more than half (51 percent) lack the skills to operate digital factories. The more successful manufacturers have advanced talent strategies in place to digitally enable the workforce of the future.
Cracking the Code on the Digital Factory, a report based on a global study of 450 manufacturers, found that a growing skills gap is one of their biggest concerns – a situation that has worsened in recent years as manufacturers have transformed their operations using new technology, analytics and mobility capabilities.
Fifty-five percent of manufacturers, up from 38 percent in 2013, reported a skills gap among skilled trades laborers, who need to operate increasingly advanced digital machinery and equipment, such as 3D printers or modeling and simulation tools on the plant floor. Likewise, 60 percent of manufacturers, up from 31 percent in 2013, cited a shortage of maintenance workers skilled in the use of predictive maintenance analytics that leverage data from embedded sensors in a machine-to-machine environment.
“For manufacturers to realize the full potential value of digital factories, they need to redesign their workforce to include new manufacturing skills, such as analytical reasoning and data-driven decision support,” said Russ Rasmus, managing director, Accenture Strategy. “Developing a comprehensive talent strategy inclusive of new digital skills is an imperative for today’s manufacturers.”
Digital Factory Leaders
The research identified a small group of manufacturers (8 percent) that outperformed their peers by increasing production and profitability by more than 10 percent since 2013. These “leaders” are more likely than their peers to understand which new skills they need for future growth and success, and have a more effective strategy to attract, develop and retain this new breed of manufacturing talent.
A majority of these leaders (73 percent) more frequently reported already having the requisite digital skills, as compared to 49 percent of other manufacturers, and they were nearly 50 percent more likely to report a higher degree of visibility into what skills they needed. That has allowed most of the leaders (81 percent) to achieve greater internal workforce mobility in roles involving digital, enabling them to match employees with managers who need those skills.
Barriers to Success
While these digital factories are enabled with rapidly developing technology innovations, the technological aspect of their implementation is not the top barrier to success. Seventy-five percent of the deployment challenges cited by survey respondents are related to skills, organizational change or structure, and the talent within the organization.
Chief obstacles that hinder manufacturers’ digital adoption.
“Manufacturers must aggressively manage these non-technical barriers as they deploy their digital factory capabilities. These include the ability to create new processes, lead teams made up of workers and machines, and constantly update training programs,” said Rasmus.
by Gary Mintchell | Jun 24, 2015 | Automation, Data Management, Operations Management
There are so many ways we can go to try to understand and then to make use of the Industrial Internet of Things. As my thinking coalesces I’ve come to the conclusion that the IIoT is a tool. It is a tool to be used in the service of an overall manufacturing/production strategy.
In order to properly use this tool of connected devices serving real-time data, we are going to need advances in data science.
Two database types seem to dominate in manufacturing—at least as expounded by suppliers. One is a relational (SQL) database. The other type is data historian.
I remember talking to some of the tech guys at Opto 22 about exploring semi-structured and open source variants such as NoSQL. At the time they thought that SQL would be all they need. And maybe so. But that was a couple of years ago.
All that discussion introduces an important podcast I just listened to. I subscribe to the O’Reilly Radar podcasts on iTunes. They’ve been cranking out about one per week—usually to promote an O’Reilly book or O’Reilly conference.
Data Science
Michael Stonebraker was awarded the 2014 ACM Turing Award for fundamental contributions to the concepts and practices underlying modern database systems. In this podcast, he discusses the future of data science and the importance—and difficulty—of data curation.
[Notes from the O’Reilly Website]
One size does not fit all
Stonebraker notes that since about 2000, everyone has realized they need a database system, across markets and across industries. “Now, it’s everybody who’s got a big data problem,” he says. “The business data processing solution simply doesn’t fit all of these other marketplaces.” Stonebraker talks about the future of data science — and data scientists — and the tools and skill sets that are going to be required:
It’s all going to move to data science as soon as enough data scientists get trained by our universities to do this stuff. It’s fairly clear to me that you’re probably not going to retread a business analyst to be a data scientist because you’ve got to know statistics, you’ve got to know machine learning. You’ve got to know what regression means, what Naïve Bayes means, what k-Nearest Neighbors means. It’s all statistics.
All of that stuff turns out to be defined on arrays. It’s not defined on tables. The tools of future data scientists are going to be array-based tools. Those may live on top of relational database systems. They may live on top of an array database system, or perhaps something else. It’s completely open.
Getting meaning out of unstructured data
Gathering, processing, and analyzing unstructured data presents unique challenges. Stonebraker says the problem really is with semi-structured data, and that “relational database systems are doing just fine with that”:
When you say unstructured data, you mean one of two things. You either mean text or you mean semi-structured data. Mostly, the NoSQL guys are talking about semi-structured data. When you say unstructured data, I think text. … Everybody who’s trying to get meaning out of text has an application-specific parser because they’re not interested in general natural language processing. They’re interested in specific kinds of things. They’re all turning that into semi-structured data. The real problem is on semi-structured data. Text is converted to semi-structured data. … I think relational database systems are doing just fine on that. … Most any database system is happy to ingest that stuff. I don’t see that being a hard problem.
Data curation at scale
Data curation, on the other hand, is “the 800-pound gorilla in the corner,” says Stonebraker. “You can solve your volume problem with money. You can solve your velocity problem with money. Curation is just plain hard.” The traditional solution of extract, transform, and load (ETL) works for 10, 20, or 30 data sources, he says, but it doesn’t work for 500. To curate data at scale, you need automation and a human domain expert. Stonebraker explains:
If you want to do it at scale — 100s, to 1000s, to 10,000s — you cannot do it by manually sending a programmer out to look. You’ve got to pick the low-hanging fruit automatically, otherwise you’ll never get there; it’s just too expensive. Any product that wants to do it at scale has got to apply machine learning and statistics to make the easy decisions automatically.
The second thing it has to do is, go back to ETL. You send a programmer out to understand the data source. In the case of Novartis, some of the data they have is genomic data. Your programmer sees an ICU 50 and an ICE 50, those are genetic terms. He has no clue whether they’re the same thing or different things. You’re asking him to clean data where he has no clue what the data means. The cleaning has to be done by what we could call the business owner, somebody who understands the data, and not by an IT guy. … You need domain knowledge to do the cleaning — pick the low-hanging fruit automatically and when you can’t do that, ask a domain expert, who invariably is not a programmer. Ask a human domain expert. Those are the two things you’ve got to be able to do to get stuff done at scale.
Stonebraker discusses the problem of curating data at scale in more detail in his contributed chapter in a new free ebook, Getting Data Right.
