Last week my team hosted an exciting event at the Four Seasons in Houston, TX progressing our efforts in this vertical. It was an event that mixed users, partners and customers plus the many faces of Hitachi. Our aim was two pronged:
- Be inspired through the continued exploration of new challenges from the industry, and
- Validate areas we're already progressing, and adjusting based upon user feedback.
Doug Gibson and Matt Hall (Agile Geoscience) kicked us off by discussing the state of the industry and various challenges with managing and processing Seismic data. It was quite inspiring and certainly revealing to hear where the industry is investing across Upstream, Midstream and Downstream -- the meat, Upstream used to be king, but investments are moving to both Midstream and Downstream. Matt expressed his passions about literally seeing the geological progression of the Earth through Seismic Data. What an infectious and grand meme!
More generally, I believe that our event can be seen as "coming out party" for works we began several years ago -- you'll continue to hear more from us as we work our execution path. Further, being inspired by one Matt Hall we ran a series of un-sessions resulting in valuable interactions.
The Edge or Cloud?
In one of the un-sessions, Doug and Ravi (Hitachi Research in Santa Clara) facilitated a discussion about shifting some part of analytics to the edge for faster and more complete decision making. There are many reasons for this and I think that the three most significant are narrow transmission rates, large data (as in velocity, volume and variety), and tight decision making schedules. Even though some processes (especially geologic ones) may take weeks, months or years to conclude when urgency matters a round trip to a centralized cloud fails! Specifically, HSE (Health, Safety and Environment) related matters, plus matters related to production of both oil and gas mandate rapid analysis and decision making. Maybe a better way to say this is through numerical "orders of magnitude" -- specific details are anonymized to "protect the innocent."
- Last mile wireless networks are being modernized in places like the Permian Basin with links moving from satellite (think Kbps) to 10Mbps using 4G/LTE or unlicensed spectrum. Even these modernized networks may buckle when faced with terabytes and petabytes of data on the edge.
- Sensing systems from companies, like FOTECH, are capable of producing multiples of terabytes per day, which join a variety of other emerging and very mature sensing platforms. Further digital cameras are also present to protect safety and guard against theft. This means that the full set of Big Data categories (volume, velocity and variety) exists on the edge.
- In the case of Seismic exploration systems, used to acquire data, designs include "converged-like" systems placed in ISO containers to capture and format Seismic Data potentially up to the scale of 10s of petabytes of data. Because of the remote locations these exploration systems operate in there is a serious lack of bandwidth to move data from edge to core over networks. Therefore, services companies literally ship the data from edge to core on tape, optical or ruggedized magnetic storage devices.
- Operators of brown-field factories with thousands of events and tens of "red alarms" per day desire to operate more optimally. However, low bit rate networks and little to no storage in the factory, to capture the data for analysis, suggest something more fundamental is needed on the edge before basic analysis of current operations can start.
This certainly gets me to think that while the public cloud providers are trying to get all of these data into their platforms there are some hard realities to cope with. Maybe a better way classify this problem is as trying to squeeze an elephant through a straw! However, many of the virtues of cloud are desirable so what we can we do?
Progressing Cloud to the Edge
Certainly the faces of Hitachi have (industry) optimized solutions in the market already that enrich data on the edge, analyze + process to skinny down edge data, and business advisory systems capable of improving edge related processes. However, my conclusion from last week is that resolutions to these complex problems are less about what kind of widget you bring to the table and more about how you approach solving a problem. This is indeed the spirit of the Hitachi Insight Group's Lumada Platform because it includes methods to engage users, ecosystems and brings tools to the table as appropriate. I was inspired to revisit problem solving (not product selling) because Matt Hall said, "I was pleased to see that the Hitachi folks were beginning to honestly understand the scope of the problem" as we closed our summit.
Is O&G the poster child for Edge Cloud? It seems that given the challenges uncovered during our summit plus other industry interactions the likely answer is yes. Perhaps the why is self evident because processing on the edge, purpose building for the industry and mixing in cloud design patterns is obvious as stacks are modernized. It is the "how" part I believe deserves attention. Using Matt's quote, from the last paragraph, guides us on how to push cloud principals to the edge. Essentially, for this industry we must pursue "old fashioned" and sometimes face-to-face interactions with people that engage in various parts of the O&G ecosystem like geologists, drilling engineers, geophysicists, and so on. Given these interactions which problems to solve, their scope and depth become more obvious and even compelling. It is then when we draft execution plans and make them real that we will resolve to build the cloud at the edge. However, if we sit in a central location, read and imagine these problems we won't develop sufficient understanding and empathy to really do our best. So, again yes Oil and Gas will engender edge clouds, but it is the adventure of understanding user journeys that guides us on which problems matter.
- Top Banner Picture - Author: Stig Nygaard, URL: Oil rig | Somewhere in the North Sea... | Stig Nygaard | Flickr, License: Creative Commons
- Seismic Image - Relinked from Burning the surface onto the subsurface — Agile, and from the the USGS data repository.