- Integrated Systems
- About Us
- Integrated Systems
- About Us
What’s next for IoT: Artificial Intelligence at the Edge | The Element Podcast - E07
Are your IT resources are increasingly supporting internet-of-things sensors and networking in the field or on the manufacturing floor? We call this "moving to the Edge" with the Edge being the place where these new data-spewing sensors live.
You may already be dealing with the new challenge of this IoT revolution:
- How do you get all this new data processed in the data center, or even transported back in a timely manner?
- What to do with all that new data from these sensors?
- How do I manage, monitor and secure the all those new sensors?
It’s too slow and expensive to send to the cloud or the datacenter and back to make adjustments on the fly. Having trouble keeping up with all this new complexity? You’re not alone. The reality is that you can't straght-line scale existing IT solutions and human analysis silos to address the opportunity buried in this underlying data.
This is where AI methoologies and skillsets need to be put to work. You need systems to be self-monitoring, computing power at the sensor or sensor node level and your network needs to envision how "right-sized" pools of data return to the mothership to be analzyed.
Here at HPE, we have already responded to this need from our customers and building core competency in managing IT resources "from the edge to the cloud". In this way, pathways are created for customers and the IT industry to begin making these big leaps.
To talk more about the intersection of AI and Iot strategies, we brought in two our two favorite thought leaders in both areas:
- Christina Skarpathiotaki, Data Science Consultant in Worldwide CoE for AI, Data and Emerging Technology at HPE
- Quentin Jones, IoT Practice from Accenture's Manufacturing, oil & gas – Industrial division
Here's a preview of the discussion (watch the entire epsiode below):
Quentin ground us in what is happening in the oil & gas sector already when it comes to IoT and where the challenge lies:
So yes I'll take it from an oil and gas scenario because I live in Houston been in oil and gas for about 20 years all the way from production up through the backhaul. So you know, sensing, uh know, instrumentation is nothing new in oil and gas. We've been doing it for years. The problem is is we've never really done anything with that data.We've just collected data for data sake as well as the aging infrastructure of radio networks to get that data back has to be a fundamental shift.
Quentin explains the solution Accenture is building for customers with HPE is a solution to this gap:
And so that's why we're looking at deploying at the edge and having a universal platform...for our customers. As well as once you collect the data, that's only half the battle. You need the outcome. And so that's where we are deploying machine learning and A.I. at the edge at the process edge where all the action is happening to make those split decisions...
Going deeper into where the IT infrastructure needs to enable insights-driven, he highlights the need to locate processing power at the "edge" and not brind it all back to the traidtional data center:
So no longer should that data sit in a data warehouse. We want to do it at the at the edge right in the computing power as well as you know there has always been proprietary companies that have kept companies locked down and not getting the full value out of that data. So that's where again we need that processing power to do the data ingestion, the data analyzation, and then spit out an outcome and eventually get to the point where it's learning from everything it's done in the past and making that automatic autonomous recommendation
Easier said than done, right? Definitely. Here's some #realtalk from Christina on enabling the benefits of AI:
Yeah. So actually artificial intelligence is actually a buzzword that everyone talks about but actually artificial intelligence it's not something new. It's something that exists already about enterprises are starting now to to get on board. So we're talking with our customers these days and they're really excited. What are the different opportunities out there? How they can benefit from the different datasets combine like data coming from the different data sources. They don't have the capabilities anymore. It’s so much data that they cannot anymore contextualize and analyze.
So part of this is bringing the data scientists and the IT teams together and begin working in tandem. Here's Quentin again:
So you know it comes around. So a lot of times what happens is you know we come to conferences and they go oh we need an AI strategy or we need a digital or die strategy. And. And nobody knows what to do with that. So that's definitely where you need to bring these teams like we talked about together and figure out how are you going to fundamentally shift what you're doing today in order to better EHS records or quicker to production time or you know a better product. Building a better widget at the end of the day.
From here, we dive deeper into the ecosystem, technology and mindsets that are going to be part of this fundamental shift to connect what's going on at the edge with the central management and control functions in a organization.
See the entire discussion below and leave a comment on where your enterprise is feeling the pain, or reaping the rewards of arriving first to this new way of growth hacking insights from your IoT.
Watch the discussion on YouTube
Hewlett Packard Enterprise