Boosting energy demand forecast accuracy by 24% for thousands of NYC buildings

Boosting energy demand forecast accuracy by 24% for thousands of NYC buildings

OneSix developed a deep learning-powered forecasting engine that improved demand forecast accuracy by 24% and automated long-range, hourly projections across 1,000+ buildings to support energy program participation.
Data Science
AI & Machine Learning
Forecasting

Overview

Manual energy forecasting process limited accuracy and scalability

An energy management firm needed to forecast electricity demand across approximately 1,000 buildings in New York to support demand-response and energy-efficiency programs such as NYISO’s Special Case Response.

The existing process was slow, manual, and error-prone—relying on spreadsheets and rough estimations to determine seasonal commitments. The forecasts had to be both granular (hourly) and long-range (up to six months), while accounting for complex, interacting seasonal patterns across a wide variety of building types.

Our Solution

AI-powered forecasting engine enables smarter, scalable planning

OneSix developed a forecasting and simulation engine powered by deep learning. The solution included:

The system was built using PyTorch and integrated with Torchcast to manage temporal data modeling and training workflows.

Results

Improved accuracy, automation, and program impact

Ready to unlock the full potential of data and AI?

Book a free consultation to learn how OneSix can help drive meaningful business outcomes.