PTC LiveWorx 2015 concluded earlier this week. PTC brought together a good mix of users and prospects, technology vendors and systems integrators, analysts and the press, in a series of presentations and discussions that created much enthusiasm and probably greater expectations from PTC as it seeks to establish a leadership position in the Internet of Things (IoT) marketplace.
One of the more important announcements made during LiveWorx that did not get as much attention as it deserved was the acquisition of analytics software company ColdLight.
ColdLight’s Neuron software presumably provides “automated predictive analytics” by “using artificial intelligence and machine learning technology to automatically and continuously learn from data, discover patterns, build validated predictive models.” A brief conversation with the company’s CTO didn’t reveal much in way of artificial intelligence or deep learning. The software, which appears to be more of multivariate statistics and a hint of a neural network type learning), certainly does not demonstrate sufficient domain-specific semantic understanding to be able to fulfill some of the promises made in the press release. The video advertising Neuron as a brain-like learning machine was uploaded in 2009. Perhaps the product direction has changed in the course of the last five years.
With reported run-rate subscription revenue of less than $8 million, and no apparent experience in manufacturing, which is the industry PTC targets as the prime market for IoT, the term Predictive Analytics Leader used to describe ColdLight may be a bit difficult to justify, and the price tag of approximately $105 million may also raise some eyebrows on Wall Street.
So why do I maintain that the addition of ColdLight to the ThingWorx IoT platform is so important?
As I’ve discussed and written in the past, the business value of the IoT is not in connecting and managing devices, although these are critical to form the necessary foundation for value realization. The real value lies in the ability to shorten the latency in decision-making. The ability to create rich multidisciplinary context for better decision-making is actually realized outside the IoT platform—in PTC’s case ThingWorx—and requires a broad set of data and a highly capable analytic tool.
PTC’s IoT strategy remains horizontal. ThingWorx is a general purpose IoT platform, augmented by tools such as ColdLight that enable customers and systems integrators to build domain-specific big data analytic applications.
This last point should serve to highlight the challenges ThingWorx is going to face.
- First, as PTC readily admits, IoT applications require deep domain expertise, both technical and business. The ThingWorx platform and ColdLight analytics do not provide those “out of the box.” PTC’s response to this challenge is aggressive growth of its technology and implementation partner ecosystem, which was highly evident at LiveWorx.
- Specific to ColdLight, the company will have to prove that its technology can handle analytic and predictive tasks in a range of data types and semantics and provide deep and rich insight in a cost effective manner. At present, it appears that the Neuron software requires a significant level of professional services.
- Lastly, as PTC targets ColdLight as ThingWorx’s main analytic tool, it will no doubt have to figure out how to cohabit with mainstream enterprise BI tools such as Cognos, BusinessObjects, HANA and SAS. This is not a question of analytic capabilities. It’s conceivable that a deeply embedded ColdLight will offer better analytic using ThingWorx generated data, but high fidelity enterprise-level decision making must incorporate a broader range of multidisciplinary data, much of which is stored in the enterprise ERP system, out of the reach of the IoT platform.