Energy Usage Forecasting System
The Challenge
Client manually estimated daily electricity usage of a manufacturing site for the following day. Any errors in the estimate (compared to actual usage) resulted in significant cost penalties for the client.
The estimates were time-consuming, inconsistent, and vulnerable to changing external factors like the weather.
The Solution
Nimble Data built an automated forecasting tool to improve prediction accuracy and reduce cost exposure:
Ingested planned production data from SAP and weather forecast data via a secure API
Developed a machine learning model to forecast site electricity usage based on production and weather forecasts
Delivered results through SharePoint: a simple and familiar interface for the operations teams
(Note: Example visuals shown; actual client data and interfaces are confidential.)
The Impact
Significantly reduced forecasting errors, leading to lower penalty costs
Freed up operational teams time from the manual estimate processes
Increased confidence in daily site planning and cost management
“Emma has been a real asset to our team. She brings a strong background in manufacturing and engineering, which means she can quickly understand the technical context and get to the heart of the problem. What sets her apart is her ability to keep the business case front and centre- always focused on delivering tangible value. She’s highly self-sufficient, proactive, and requires minimal oversight, which makes collaborating with her refreshingly easy and efficient.”
