DAPCOR (Digital Assistant for Plant COntrol Room) is a virtual assistant tool written in Python and uses LLM to reduce the barrier to information access for process plant personnel by accepting voice-based queries in natural language to analyze process data.
Users can ask their queries in their natural language, and DAPCOR will fetch and process relevant data as needed and provide appropriate answer. The idea is that if you are in a plant control room and you want to know something, you should not have to go to your computer and open dashboards or trending tools just to get some quick information – DAPCOR should be able to do that for you. (DAPCOR is not meant to replace the dashboards and trending tools, but make their results more accessible to control room personnel.)
Current Features
— listens to voice commands; ‘wakes up’ upon hearing ‘OK Digital Assistant’
— fetches data from SQL query and verbally answers the user
— plots appropriate graphs based on the query
Upcoming Features
— ability to perform automated advanced data analysis using Python to adequately respond to user queries
— ability to accept follow-up queries
— RAG abilities to utilize enterprise documents while framing response
— plots appropriate graphs based on the query
Check out the GitHub repository for more details.