Case Study: Halliburton Energy Services achieves safer explosive detonation detection and 100× faster development with MathWorks (MATLAB & Neural Networks)

Halliburton Makes Oil Exploration Safer Using MATLAB and Neural Networks

Halliburton Energy Services needed a reliable way to confirm that explosive charges had detonated in deep wellbores despite heavy ambient noise from pumps and generators. Halliburton engineer Roger Schultz used MathWorks tools — primarily MATLAB with Deep Learning Toolbox and MATLAB Compiler — to prototype and package a field‑ready filter and application to separate impulsive detonation signals from repetitive machinery noise.

Using MATLAB and the Deep Learning Toolbox, Schultz built an adaptive, predictive nonlinear neural‑network filter and compiled it into a standalone application with MATLAB Compiler so teams can record, filter, and play back signals at the well site. The MathWorks solution delivered production‑grade accuracy and authentic desktop simulation, accelerated development (about 100× faster than Visual Basic or C), and led Halliburton to pursue patent protection for the technology.

Halliburton Energy Services

Roger Schultz

Research Engineer