Root Cause Analytics

Why did we apply low/no-code layer to Fabrik?

In the early days of Fabrik, we did a lot of programming - customers asked for specific workflows and we programmed them. They asked us to make changes, we went back to our code and changed. They asked for some more changes, and we wrote some more code to correct. This pulled our developers from product roadmap to cosmetic fixes over and over again. We realised quite soon that we needed to decouple the platform from the experienced developer and the low/no-code layer was added as a necessity. You can change "experienced developer" to an experienced person at your workplace and you realise where low/no-code layer combined with visualisation is valuable in your context.

Armed with this tool, we have translated tacit knowledge in subject matter experts' minds into guided workflows for applications ranging from training, simulations, day-to-day operations, and design reviews. In each of these scenarios, there are certain conditions that determine the course of action, and the key is enabling users to define them with a few clicks without the need for advanced degrees, extensive work experience, and costly computers.

By giving this power to people closest to the problem (or assets and processes), they contribute to creating accurate root cause analytics for a given asset/process in a closed system.

How can you apply Fabrik's Root Cause Analytics to your systems?

Step 1 - aggregate:

Root cause analytics is a set of rules applied on different parameters and specific actions are recommended. The parameters come from different systems like:

1. Sensor data from IoT, SCADA, PLCs, and inference algorithms.
2. Standard operating procedures as PDFs
3. Data from historians
4. Tribal knowledge accrued over decades
5. Design and performance data

A combination of two or more from above list is good to get root cause analytics applied for any asset.

Step 2 - compose:

Upload or integrate above parameters into Fabrik through the authoring studio (drag-and-drop interface) and define workflows based on one or more conditions. These can be one or many procedures interconnected with each other for any asset/system. Define conditional workflows based on one or more of these multi-dimensional sources and trigger individual/specific activities based on the conditions of the system.

Step 3 - learn:

Capture usage feedback from the field to add more details, more data points, nuanced conditions, and modify the procedures on-the-fly to reflect best practices.

Step 4 - deeper integration:

PDFs and powerpoints have been the standard modes of knowledge transfer between diverse teams and across generations. As the rank-and-file transition to 3D workflows with built-in root cause analytics, there is a lot of data generated on user's ability to follow instructions, the accuracy of root cause identification, and time to issue resolution. Connecting the response from the field to improve root cause analytics at asset, process, and facility levels improve overall operative effectiveness.

Achieve high-quality interventions in real-time and through simulations

Aside from real-time learning and implementation, the ability to use virtual reality enables us to run trainings, simulations, and emergency responses where no amount of presentations and documentation work. Real-world data and learning enables us to capture the complexity of the real world reasonably well for all scenarios and prepare for it. Given the challenges of coordinating with diverse teams take time off their day for simulations (think disaster response or grounding aircrafts for training), using Fabrik's root cause analytics allows teams to access and run key aspects of the digital twins on their personal devices at a fraction of what is costs today to run the same simulations.