Bitron boosts EV charging station efficiency with Oracle AI solutions
“Oracle AI Vector Search in Autonomous AI Database simplified all our company data and information and allowed us to apply conversational GenAI to dramatically improve field service resolution in a secure and scalable way.”
Bitron Electronics designs and manufactures electronic components and systems for the household appliance, automotive, energy, and other sectors. As part of its monitoring of more than 200 electric vehicle charging stations, pulling together operational data in fault tickets, technical manuals, and other sources spread across multiple platforms, Bitron found it challenging to manage malfunctions and ensuing claims. Growing demand for rapid diagnosis and responsiveness required an integrated approach. Using Oracle AI Database on Oracle Autonomous Database, along with Oracle Cloud Infrastructure (OCI), Bitron created an intelligent chatbot that vectorizes searches of all data sources, creating a unified view to speed diagnosis and solution suggestions for its charging stations.
Why Bitron Electronics Chose Oracle
Bitron leveraged its existing Oracle Autonomous Data Warehouse on OCI equipped with AI Vector Search to search all structured and unstructured data types without moving data. The converged Oracle AI Database 26ai has native integration with various large language models (LLMs) and uses retrieval-augmented generation (RAG). Combined with OCI Generative AI and its flexibility of using Cohere’s Command R and Meta’s Llama LLMs, the Oracle AI solutions enable Bitron to analyze error codes and deliver accurate and coherent recommendations during the production process.
Results
The conversational chatbot assistant has improved the way Bitron manages malfunctions and claims, answering 85% of operator questions. Using centralized historical data and technical documentation, operators can query the assistant based on error codes and telemetry from the charging stations to diagnose issues 5X faster than before. The ability to suggest targeted interventions has reduced charging station downtime and operational costs.
In addition to improving diagnostic accuracy, the combined power of LLMs and RAG with an integrated Bitron knowledge base has reduced the amount of time spent on both manufacturing as well as monitoring and maintaining the charging stations by tens of hours per month.
The solution is designed to be scalable, allowing the company to extend these AI capabilities to other products in its portfolio. Oracle Analytics Cloud has helped Bitron reduce the time required to access all its production process data by 90%, to about 1 minute, empowering line operators and executives with a real-time dashboard. This unified view of all company data and analytics enables Bitron to optimize KPIs, reduce risk, and standardize every step of its production processes.
Partners
Bitron partnered with Technology Reply, which has specialized in the design and implementation of Oracle AI, machine learning, analytics, and security technologies for more than 25 years across a range of industries.
About the customer
Based in Italy, Bitron is a family-owned company with nearly 10,000 employees and 18 manufacturing plants worldwide.