Jun 28, 2019
| 10 min read

Spotlight Series: How Uptake's AI-Driven Platform Enables Next-Gen IIoT Success

This series highlights the key insights and lessons from our Digital Leadership series of podcasts. We spotlight the important takeaways from our interviews in an accessible format. The following insights come from Jay Allardyce, EVP Corporate Development & Partnerships for UptakeStay tuned for the full podcast interview with Jay Allardyce, in the meantime, take a look at our full library of podcasts.

What are some of the key experiences that have brought you to where you are today in your current role?

I’m a Silicon Valley native and grew up not too far from Hewlett-Packard, and it was really a lot of the influence through family members and neighbors that it all aspired over the years to go work at Hewlett-Packard, to HP Labs. I’ve always been in tech, started early in my career as a developer, and realised that was not a long-standing suit for me, but I always had this interesting view of inter-connected systems and environments. This goes back to EDI and RosettaNet and spending a lot of time configuring these environments to connect enterprise supply chain systems, and channel systems.  

I spent about 15 years in Hewlett-Packard in a variety of roles but I think the most notable for me 10 or 12 years ago was in HP Labs, helping to bring together this product offering around real-time energy management, out of thinking what it could bring to the market from a technology-datacenter perspective, to buildings and broader environment. It was the simple principle of, ‘Can you create insight from the data that you have access to, to affect the way the business works?’ That has stuck with me I think for the last 10 or 12 years in a variety of roles, with Vertica and some of the big data areas, then spending a number of years at GE.

It’s been a culmination, and the premise that we do live in a data-driven world, and we’ve seen it really explode on the B2C side of things, and I believe there’s a mountain of opportunity in the industrial and enterprise space when it comes to IoT.

Could we get your perspective on how energy differs from other industries that may have been earlier to adopt analytic technologies??

In my early exposure at Hewlett-Packard, it was really exciting to see a way to potentially change the way our consumer behavior thinks about energy use and waste. When I went into GE I was interested in going back to help that problem and advance it, and in this case of realizing that we have to live in the notion of an energy value chain, from a generation, to distribution, to consumption, realizing it’s becoming a ubiquitous network, and the rise of the prosumer was coming forward, and still is with the onset of renewables and solar, and the ability to think about battery storage and other technologies.

With that there’s going to be a massive impact to the grid, or the existing ways in which utilities have operated. So, I’ve long been fascinated with that way of understanding how you disrupt an industry in a right way and in a positive way, using data. I think partly why I was gravitated towards energy is, 1) a very hard problem to solve, 2) an industry that has a significant amount of data, and 3) an industry that knows it needs to reinvent itself, and 4) which is probably the learning coming into, or being a part of it is just it takes a long time. Change is definitely hard, and regulated or unregulated markets, and/or just the habit of the way things have operated in the past. It’s not to say utilities aren’t willing to make that change, I just think there are a number of factors that’s putting effort in the way of how quickly they adopt and transform. I like what I’m seeing, and the acceptance and acknowledgment across the CEO communities for utilities that acknowledge that this is important not simply to do, but to ensure that it’s a part of the fabric of utilities going forward.

I think what’s interesting about what you’re doing at Uptake is how the AI components, or machine learning is embedded into the platform, and into the DNA of the company. How you combine all of those elements together to drive outcomes for your customers?

To give you an example, last week we hosted about 30 different, super-effective customers in the construction space. There was everything from operators, to very different types of OEMs, and a very strong cross-section of individuals, and that very question came up. What I shared with them was, first get a sense of where people are in their journey, if we talked the notion of digital transformation, but more importantly their acceptance and willingness of leveraging AI machine learning.

Step one was just simply dispelling the myth where there’s the general concern, especially in the industrial space, that the rise of AI equals job loss. With that is the natural acknowledgement first around the fear their job might be redundant, and therefore getting people onboard to think about they are possible. And it’s a straight question to oftentimes there’s an opportunity to improve cost structures, and so that usually is the easiest low-hanging fruit, and then just start talking about how you think about the day, and how you operate.

I’ll give you the example on the construction environment, we’ve got scenarios where customers are like, ‘I constantly don’t know if I have the right bench of individuals that have the right skillset, and are available at the right time, when I need to go and fix a piece of equipment that I’ve been alerted is down. Oftentimes I’m questioned, do I have the right person, the right tool, the right part at the right time, and can I go fix the unit itself to get it back online, and continue to drive better productivity in its use, and avoid the downtime. That is in essence a way of us just simply looking at the problem set that someone has had, and say, ‘Okay, look back through that and understand what problem we’re trying to solve, and therefore ask the question of, what data do we have access to today?’

Are there notable differences in working with some of the industries, or different industries that Uptake works with?

What we found is there’s a fair bit of reusability with different data science engines and models that we’ve deployed, they’ll allow it to be able to move more quickly for the first to insight. So, in the example in the case of auto manufacturing, it is really about managing the uptime and the production of various lines that our customers have and trying to be mindful that it’s not simply looking at the downtime of a given asset, but it’s looking at the entire process itself. What we do is very similar to what we’ve done in other industries, is looking at the asset in their given setting, and not just on an individual basis, but the interconnectedness of those to understand what metric that customer’s carrying to improve. If you think about OEE, one of the biggest things is ensuring that obviously the asset availability and up-time is there, such that the production line can be as effective as it can be, on its command and planned schedule.

Then I’d say one of the things we’re working heavily on is with the US Army in the Federal space, a number of opportunities; but a lot of it has been publicly known working with the US Army around the Bradley fighting vehicles, and helping to make sure we have operational readiness for our wartime heroes and fighters, to make sure the equipment that they use is available and ready for dispatch and deployment when it matters most. In some cases, we have to be segmented in the way we work, especially with military; but the principles and mindsets of how we approach our data science model in the commercial side, helped us dramatically to glean insights and ways to evolve these models more quickly for the benefit of our customers.

How do you think about the way you turn your business to be able accelerate adoption for new customers and new use cases while still keeping that the ability to customize your technology for new situation and scenarios?

The very important thing is, certainly data scientists that are very well versed in the discipline can manage very large datasets and come up with very distinct models. But it’s not enough in the viewpoint that the expertise from an industrial perspective needs to wait in terms of understanding and interpreting the value. One of the things I value is this notion that as much as we would talk about care programming, peer reviews, and things that really help each other from a development perspective get better at what we’re ultimately releasing in the market, the same applies when you’re putting a subject matter expert who’s had 20-years in a given industry, working close to the data, and being able to interpret that.

Then the other effort of it is the constant knowledge sharing we have of what teams have seen with certain industry use cases, and challenges, and the tight-knit community of how the data science group has been established. The uptake which I think has been absolutely tremendous over the years to help evolve that, and then it’s also the way you think about the technology, and that is knowing you have very consumable capabilities, or what we call engines that can be applied to different use cases depending on the customer, the type of asset situation. But the ability to deploy that application is very consistent and repeatable, such that again we haven’t built a one-off for a given customer, and we have the ability to leverage that.

 

orange_line_1-759860-edited

Momenta Partners encompasses leading Strategic Advisory, Talent, and Investment practices. We’re the guiding hand behind leading industrials’ IoT strategies, over 200+ IoT leadership placements, and 25+ young IoT disruptors. Schedule a free consultation to learn more about our Connected Industry practice.