Digital Transformation Strategy: Moving Beyond BuzzwordsDigital Transformation Strategy: Moving Beyond Buzzwords
Vague concepts like AI and digital transformation strategy can lead industrial firms to lose track of their core problems. But the situation could be improving.
June 14, 2019
BOSTON — This year, the LiveWorx Digital Transformation Event returned to Boston. The “digital transformation” theme refers to more of a journey than a destination, as PTC Chief Executive Officer Jim Heppelmann pointed out in a keynote address. And this term encompasses a diverse set of technologies. “It’s computation, sensors, networks, AI, robotics, 3D printing, synthetic biology, AR/VR, blockchain,” said Peter Diamandis, founder and chairman of the X Prize Foundation in another keynote address at LiveWorx, only mentioning a handful of the terms fitting under the digital transformation rubric.
For a company to have successfully transformed, however, it must have a clear form of ROI to show for it. For example, Netflix has digitally transformed its business to the extent it has converted the bulk of its revenues from a subscription based on physical goods to one primarily based on streaming content. Founded in 1997 as an online video store shipping DVDs by mail, the company unveiled its streaming service a decade later. Its annual revenue in 2007 was $1.2 billion. Eleven years later, it was $15.8 billion. Netflix founder Reed Hastings was able to “jump [his business] onto an exponential growth curve” in moving “from the post office to broadband,” Diamandis said. The company also deployed an analytics engine for its streaming service intended to recommend content based on user feedback, making it a machine learning pioneer as well.
But the question of how to measure digital transformation in the industrial sector is a different animal. For one thing, there is tremendous variability in even just manufacturing. For another, a manufacturer with decades’ worth of infrastructure and business relationships often has less freedom to reinvent itself than a consumer-focused company whose products center around software. Few companies can expect to achieve a nearly fifteen-fold increase in revenue in just over a decade.
On top of that, other benchmarks often take precedence for manufacturers such as uptime, efficiency, safety and so forth.
“What I would say, is today, digital transformation ends up being a little custom for every company and therefore their metrics and roadmap are different,” Heppelmann said in a Q&A session a LiveWorx.
At present, many industrial firms rely on consultants who can help craft a custom digital transformation initiative.
PTC aims to create a new business unit known as Digital Transformation Solutions, aiming to help standardize this type of metamorphosis. “We’re trying to package this up a little more, but really start with the value prop,” Heppelmann said. Instead of pitching a smattering of technologies, PTC would begin conversations with industrial customers by asking what their biggest challenges are and what the value might be if those hurdles were resolved. PTC would then aim to recommend a set of technologies designed to address that problem, rather than recommending broad umbrella technology themes such as IoT, AR or AI from the get-go.
If it sounds like a service a management consulting might offer, that’s no accident. PTC enlisted a partner at a prominent consulting firm to lead its new Digital Transformation Solutions business.
The transition, however, isn’t wholly new for the company. “Two or three years ago, ThingWorx was primarily a development platform where people could build apps. We are moving more and more toward purpose-built solutions aimed at specific economic value,” Iain Michel, divisional vice president and general manager, ThingWorx at PTC told IoT World Today.
The company is positioning ThingWorx and its own experience working with customers as a means to help provide the majority of digital transformation technologies. “And we’re trying to say to customers: ‘Let us build the first 80%. Every single one of you is building the same first 80%. So don’t waste your time building that,’” Michel said. The customer can then build out the remainder of the project. “You build the final mile that is special sauce for you or for your industry or as a competitive differentiator,” Michel said. “So that’s where you do get 80% reusability with 20% bespoke, but to the individual customer, it feels fully bespoke.”
As for where the concept of artificial intelligence fits into digital transformation, it is a proverbial tool in the toolbox. “And I think we’re early [with AI] and when it became a buzzword, everybody said: ‘Well, you gotta do AI,’” Michel said. “And various people were sent off to go to an AI project. But AI is a means to an end, not the end itself.”
PTC, for instance, is using AI-based technologies to recognize objects and for maintenance applications where pattern recognition is possible.
“The problem with AI in a piece of machinery is that unless you can get enough data coming into it, either from the OEM, or from a number of different similar types of equipment that you have around the factory, stuff just doesn’t go wrong that often,” Michel said.
To train a machine learning model to spot an anomaly so the system can predict a future glitch can be a slow process. “You often need years’ worth of data and several failure events that are predictable,” Michel explained.
Here, OEMs with connected machinery have an advantage in that they have a broader view into how their customers are using their equipment in the field, and potentially have sufficient data to recognize what a failure event looks like and predict such events in the future.
“In the service world for sure, this predictive maintenance idea is the Holy Grail,” Michel said.
But some organizations might have more success focusing on ideal operating conditions than trying to predict anomalies. “You can get an awfully long way by knowing what the right operating conditions for a piece of equipment are,” Michel said. “And then when it pops out of peak operating conditions, you can say: ‘Hey, look, for whatever reason, it’s not operating in the optimal range. Go look at it.’ So we’ve got a lot of customers are doing that.”
Another dimension relevant to big technological themes such as AI is the broader ecosystem of vendor partners, said Justin Hester, digital transformation director at PTC. “Really, it’s a team sport. It’s about [industrial companies] moving away from saying: ‘I want AI/ML’ to say: ‘Hey, I have a challenge. Here’s the outcome I want. Let’s go find partners and solutions that get me there,’” said Hester, who was a PTC client before coming to work for the company. While it doesn’t make sense to deploy a host of digital technologies haphazardly, after identifying a core problem statement, it is a good idea to find a partner ecosystem capable of providing a “massive toolbox” to help meet the needs of the digital transformation strategy as it evolves over time.
“I would say: AI/ML is the future. It’s amazing,” Hester said. “But I don’t suggest doing an AI/ML thing for the sake of it. You have a current situation and a desired outcome. And that may or may not involve those solutions today or tomorrow, right?”
About the Author(s)
You May Also Like