Representative End User Clients
Representative Automation Clients
Representative Software Clients
IIoT Viewpoints is always on the lookout for disruptive IIoT-enabled solutions that address longstanding problems. ARC interviewed Vish Avasarala, Co-founder of Saint Software Consultants, and Kenneth Smith, General Manager, Energy at Hortonworks about how Big Data, analytics, and data management systems can help industrial companies optimize operations performance. Today we present part one of two. Part two will be published next week.
IIoT Viewpoints: We are going to be talking about IIOT and analytics and why it's so important for the industrial world to be paying attention to these groundbreaking technologies. Before we do that, Vish, if you could, give us a little background on Saint Software Consultants and then Kenneth, I'll ask you for a little background on Hortonworks, and then we'll get into a couple of questions. Go ahead, Vish.
Vish Avasarala: We are a big data startup specifically focused on industrial data and analytics. Through our big data accelerators and services, we allow enterprises to reduce time-to-value from data analytics by more than 80%. We endeavor to deliver 10x ROI to our customers within the first year of their journey with us and we have been successful in this so far. That's a quick three sentence summary.
IIoT Viewpoints: All right, cool. Obviously, people can learn more about you guys at the website and we want to get straight into our questions. Kenneth, give us the quick 2-minute Hortonworks overview.
Kenneth Smith: Absolutely. Hortonworks, is an open source data platform company based out of Silicon Valley. We've been in business now for a little over six years, publicly traded, growing tremendously. Our company helps businesses across all verticals achieve their data management and enterprise analytic initiatives through the open source technologies that we contribute to, distribute, service, and support.
We lean tremendously on partners like Saint Software ultimately to deliver time-to-value through advanced analytics. We provide a platform for companies to store large volumes of historical data to then build predictive models that can be applied to real-time data, and it is companies like Saint Software that are ultimately delivering new applications and advanced analytics on the data once it's ingested and stored in our platform.
IIoT Viewpoints: Among our audience are the professionals that are managing the industrial control systems and networks. Now that we're seeing data and analytics start to touch this world, why do you think industrial companies should pay more attention to data and analytics?
Vish Avasarala: In industrial companies, if you look at the past few decades, manufacturing techniques have converged and no longer provide differentiating capability and process efficiencies. They have reached a point where you are essentially going against the laws of physics now. The cleanest way to impact bottom line, or top line for that matter, is to attack inefficiencies and optimize processes for industrial companies. An effective way to do this is to use big data analytics. Time and again, we have seen that we are able to generate value across the spectrum of a company's operations using these ideas.
IIoT Viewpoints: Okay. Give me a couple of examples, Vish. Where can you think of some possible applications of big data and analytics for industrial companies?
Vish Avasarala: If you look at a typical oil refinery, there may be 2,000 to 3,000 control loops that are feeding data to your Historian or SCADA system. Surveys have found that, more than 30% of these control loops, could be running on manual. Essentially you bought this very expensive equipment and you're not using it. That could translate to anywhere between 1.5 to 2.5 million underutilized assets per refinery. Many of these control loops are also not currently tuned. Even at the lowest level, understanding and optimizing your control strategy has a tremendous impact on operational efficiencies.
If you think a little bit more upstream, using control system data combined with your maintenance system data, can reduce non-productive time and maintenance costs across various asset types. For example, in my own personal experience, we have realized, on a conservative estimate, about 10 to 15% reduction in unplanned downtime using analytics. At a more holistic level, we can bring operational and business system data together for cross-functional analytics - like a more optimized logistics strategy. End to end, the spectrum of impact is very wide.
IIoT Viewpoints: Okay. All right, Vish, thank you. Kenneth, anything that you'd like to add regarding open source just to follow up on Vish's comments?
Kenneth Smith: One of the big use-cases we're seeing, is helping industrial companies unlock siloed data sets from both operational technology and information technology systems, enabling OT/IT convergence. Open source technologies like Hadoop provide low cost storage and parallel processing across both OT and IT data sets to deliver insights that essentially were very difficult to achieve before using existing technologies.
I come from the operational technology world where I worked for a SCADA platform vendor for a number of years. Our platform was the market leader in upstream oil and gas and had a very loyal following. However, one of our customers’ biggest complaints was once the data landed in our platform’s historian it essentially became dark data. It was basically locked down and our customers had a very difficult time accessing data to run simple reports, much less building advanced analytic applications. It was the definition of a data silo.
We’re seeing a lot of interest from industrial companies with those same challenges wanting to leverage open source technologies to consolidate data from their distributed DCS and SCADA systems into an open enterprise data historian. By centralizing all of their process data into a single platform using an open source IIoT architecture, they’re able to democratize industrial data analysis and have a single-view of operations.
IIoT Viewpoints: Vish, we know data science has been really successful in the digital world but in your opinion, are the industrials lagging?
Vish Avasarala: Traditionally yes. For example, if you look at companies like eBay or Google, analytics projects are being translated from a concept to proof of concept to A/B testing to deployment in a matter of weeks. From an idea to something that the customers can touch and feel in the matter of a few weeks. If you look at industrial companies, historically many of these ambitious data analytics projects have reached only the maturity of proof of concept. The reason for that is because with industrial work flows are complex and non-digital. Change management is a huge issue.
The fundamental nature of data is also very different. If you're talking to industrial systems, it essentially means you have tens or dozens of OEM players with different interfaces. Then if you are able to even talk to the systems, the data that you're getting is time series data. Time series analytics is different and time series persistence is different…they are not usually amenable to standard approaches that IT is familiar with, like relational databases. A lot of these challenges have existed and data science has not had the impact that it had in digital companies.
Kenneth Smith: Yes, industrial companies have historically had a laggard mentality to adopting new technologies and methodologies like data science. However, in oil and gas, for example, pricing pressures have ultimately decimated the work force. We've had hundreds of thousands of layoffs and most companies have been forced to become more digitally driven to do more with less. Many of platforms they’re accustomed to using aren't helping them achieve those goals, so they're adopting data science and technologies like open source to develop applications to help make their work force more efficient.
IIoT ViewPoints: Join us next week for Part 2 of this discussion, where we'll explore what's really changed to make the use of these technologies so more compelling for industrial operations.
About your Guest Bloggers:
Viswanath Avasarala is the founder and CEO of Saint Software Consultants. Saint Software specializes in big data services that enable enterprises to generate value at scale from data and analytics. Vish started his professional career at GE Research, developing machine learning applications for industrial equipment. Later, he co-founded Sqord, a social health focused IOT platform for kids. Sqord has 100,000+ paying customers and multiple national awards. Vish’s experience also includes leadership roles at SAS, where he led unstructured analytics product development and at Schlumberger IT where he started and led big data and advanced analytics programs. Vish Avasarala has a Ph.D in Information Sciences and Technology from Pennsylvania State University, where he conducted research in multiple sensor data fusion and an undergraduate degree in Chemical Engineering from Indian Institute of Technology, Madras.
Kenneth Smith - General Manager, Energy, Hortonworks . As the General Manager, Energy, Kenneth is responsible for establishing and leading the execution of a winning go-to-market strategy for Hortonworks in the energy industry. Responsibilities include proactively working with customers, Hortonworks sales & solution management, partners, VARs, and prospects to identify new solutions and industry specific product requirements for the energy market. He also ensures appropriate energy industry expertise support is provided and leveraged during the sales and solution implementation cycle, develops successful thought-leadership and solution strategies to drive adoption of Hortonworks products and services in the segment. Prior to joining Hortonworks, Kenneth was the Executive Director, Oil & Gas for Bit Stew Systems, and previously was the Software Sales Manager, North America for Weatherford International.
IIoT Viewpoints: We are going to be talking about IIOT and analytics and why it's so important for the industrial world to be paying attention to these groundbreaking technologies. Before we do that, Vish, if you could, give us a little background on Saint Software Consultants and then Kenneth, I'll ask you for a little background on Hortonworks, and then we'll get into a couple of questions. Go ahead, Vish.
Vish Avasarala: We are a big data startup specifically focused on industrial data and analytics. Through our big data accelerators and services, we allow enterprises to reduce time-to-value from data analytics by more than 80%. We endeavor to deliver 10x ROI to our customers within the first year of their journey with us and we have been successful in this so far. That's a quick three sentence summary.
IIoT Viewpoints: All right, cool. Obviously, people can learn more about you guys at the website and we want to get straight into our questions. Kenneth, give us the quick 2-minute Hortonworks overview.
Kenneth Smith: Absolutely. Hortonworks, is an open source data platform company based out of Silicon Valley. We've been in business now for a little over six years, publicly traded, growing tremendously. Our company helps businesses across all verticals achieve their data management and enterprise analytic initiatives through the open source technologies that we contribute to, distribute, service, and support.
We lean tremendously on partners like Saint Software ultimately to deliver time-to-value through advanced analytics. We provide a platform for companies to store large volumes of historical data to then build predictive models that can be applied to real-time data, and it is companies like Saint Software that are ultimately delivering new applications and advanced analytics on the data once it's ingested and stored in our platform.
IIoT Viewpoints: Among our audience are the professionals that are managing the industrial control systems and networks. Now that we're seeing data and analytics start to touch this world, why do you think industrial companies should pay more attention to data and analytics?
Vish Avasarala: In industrial companies, if you look at the past few decades, manufacturing techniques have converged and no longer provide differentiating capability and process efficiencies. They have reached a point where you are essentially going against the laws of physics now. The cleanest way to impact bottom line, or top line for that matter, is to attack inefficiencies and optimize processes for industrial companies. An effective way to do this is to use big data analytics. Time and again, we have seen that we are able to generate value across the spectrum of a company's operations using these ideas.
IIoT Viewpoints: Okay. Give me a couple of examples, Vish. Where can you think of some possible applications of big data and analytics for industrial companies?
Vish Avasarala: If you look at a typical oil refinery, there may be 2,000 to 3,000 control loops that are feeding data to your Historian or SCADA system. Surveys have found that, more than 30% of these control loops, could be running on manual. Essentially you bought this very expensive equipment and you're not using it. That could translate to anywhere between 1.5 to 2.5 million underutilized assets per refinery. Many of these control loops are also not currently tuned. Even at the lowest level, understanding and optimizing your control strategy has a tremendous impact on operational efficiencies.
If you think a little bit more upstream, using control system data combined with your maintenance system data, can reduce non-productive time and maintenance costs across various asset types. For example, in my own personal experience, we have realized, on a conservative estimate, about 10 to 15% reduction in unplanned downtime using analytics. At a more holistic level, we can bring operational and business system data together for cross-functional analytics - like a more optimized logistics strategy. End to end, the spectrum of impact is very wide.
IIoT Viewpoints: Okay. All right, Vish, thank you. Kenneth, anything that you'd like to add regarding open source just to follow up on Vish's comments?
Kenneth Smith: One of the big use-cases we're seeing, is helping industrial companies unlock siloed data sets from both operational technology and information technology systems, enabling OT/IT convergence. Open source technologies like Hadoop provide low cost storage and parallel processing across both OT and IT data sets to deliver insights that essentially were very difficult to achieve before using existing technologies.
I come from the operational technology world where I worked for a SCADA platform vendor for a number of years. Our platform was the market leader in upstream oil and gas and had a very loyal following. However, one of our customers’ biggest complaints was once the data landed in our platform’s historian it essentially became dark data. It was basically locked down and our customers had a very difficult time accessing data to run simple reports, much less building advanced analytic applications. It was the definition of a data silo.
We’re seeing a lot of interest from industrial companies with those same challenges wanting to leverage open source technologies to consolidate data from their distributed DCS and SCADA systems into an open enterprise data historian. By centralizing all of their process data into a single platform using an open source IIoT architecture, they’re able to democratize industrial data analysis and have a single-view of operations.
IIoT Viewpoints: Vish, we know data science has been really successful in the digital world but in your opinion, are the industrials lagging?
Vish Avasarala: Traditionally yes. For example, if you look at companies like eBay or Google, analytics projects are being translated from a concept to proof of concept to A/B testing to deployment in a matter of weeks. From an idea to something that the customers can touch and feel in the matter of a few weeks. If you look at industrial companies, historically many of these ambitious data analytics projects have reached only the maturity of proof of concept. The reason for that is because with industrial work flows are complex and non-digital. Change management is a huge issue.
The fundamental nature of data is also very different. If you're talking to industrial systems, it essentially means you have tens or dozens of OEM players with different interfaces. Then if you are able to even talk to the systems, the data that you're getting is time series data. Time series analytics is different and time series persistence is different…they are not usually amenable to standard approaches that IT is familiar with, like relational databases. A lot of these challenges have existed and data science has not had the impact that it had in digital companies.
Kenneth Smith: Yes, industrial companies have historically had a laggard mentality to adopting new technologies and methodologies like data science. However, in oil and gas, for example, pricing pressures have ultimately decimated the work force. We've had hundreds of thousands of layoffs and most companies have been forced to become more digitally driven to do more with less. Many of platforms they’re accustomed to using aren't helping them achieve those goals, so they're adopting data science and technologies like open source to develop applications to help make their work force more efficient.
IIoT ViewPoints: Join us next week for Part 2 of this discussion, where we'll explore what's really changed to make the use of these technologies so more compelling for industrial operations.
About your Guest Bloggers:
Viswanath Avasarala is the founder and CEO of Saint Software Consultants. Saint Software specializes in big data services that enable enterprises to generate value at scale from data and analytics. Vish started his professional career at GE Research, developing machine learning applications for industrial equipment. Later, he co-founded Sqord, a social health focused IOT platform for kids. Sqord has 100,000+ paying customers and multiple national awards. Vish’s experience also includes leadership roles at SAS, where he led unstructured analytics product development and at Schlumberger IT where he started and led big data and advanced analytics programs. Vish Avasarala has a Ph.D in Information Sciences and Technology from Pennsylvania State University, where he conducted research in multiple sensor data fusion and an undergraduate degree in Chemical Engineering from Indian Institute of Technology, Madras.
Kenneth Smith - General Manager, Energy, Hortonworks . As the General Manager, Energy, Kenneth is responsible for establishing and leading the execution of a winning go-to-market strategy for Hortonworks in the energy industry. Responsibilities include proactively working with customers, Hortonworks sales & solution management, partners, VARs, and prospects to identify new solutions and industry specific product requirements for the energy market. He also ensures appropriate energy industry expertise support is provided and leveraged during the sales and solution implementation cycle, develops successful thought-leadership and solution strategies to drive adoption of Hortonworks products and services in the segment. Prior to joining Hortonworks, Kenneth was the Executive Director, Oil & Gas for Bit Stew Systems, and previously was the Software Sales Manager, North America for Weatherford International.