ThingWorx Analytics - questions in 2026
Hello,
I did some research on ThingWorx Analytics, but most information I found is quite outdated — the latest updates seem to be from around 2020. In our company we use the ThingWorx platform and we would like to implement some AI-based enhancements for our condition monitoring.
I still have a few unanswered questions:
- Are there any recent updates regarding neural networks or the overall capabilities of this product?
- What are the requirements for implementing it into our existing ThingWorx platform? (For example, a specific platform version or any additional components needed?)
- How much information can it use when building a high‑quality neural network?
For example: I have raw machine data such as position, temperature, vibration, RPM etc. — the standard numerical data.
But I also want to use maintenance data (interval maintenance) and fault data (annotated in historical datasets for learn the neural network how data looks like when these events occur). These types of data can also be non‑numerical, such as a list of possible faults or machine parts replacement records that could affect machine behavior. There are many variables that can influence the machine (surrounding environment - summer/ winter temperatures can influence machine).
.
So my main question is: How does the neural network handle these non‑standard, non‑numerical data types?
Thank you.

