BRAINBOX PRIME

Autonomous Fatigue Life Assessment

The system combines machine learning and physics to prescribe current health and remaining useful life of an attack vessel under various loading and operating conditions

​

• Estimating loads (external and internal) effects on the structure

• Identify and locate fatigue hotspots (localization)

• Predicting structural damage caused by loads

• Predictive damage growth in operational conditions (inbuilt fractography)

• Measuring hydrodynamics load

• estimating the future performance of the structure by consideting the condition of the structural components (stiffeners, member elements)

BRAINBOX SPICA

Cracks and Defects Identification of Gas Cylinders

Brainbox SPICA the delivery model of Prime has been implemented in Bharat Gas, to identify cracks and other anomalies. 

​

• Crack Identification

• Leakage Detection

• Welding joint Failures

• Cylinder Bottom Checks

• Valve Leakage Detection

• Post rejection Analysis

BRAINBOX KNOWHERE

Autonomous Feeder Boat Control System

Its a Novel project, which includes design of feeder boat with AI based control algorithm, which provides the feed to shrimp effectively. Brainbox Knowhere is been used to resolve this complex operation  

​

• AI Based boat design (state, weight, heading, feed

• Influencer mapping (environmental loading), with wave and current

• Planning and control

• Feed disbursement based on requirement

• Propeller Capability and Thrust Control

• Next Best Action for Anticipating deeding Distance, Heading Area, Feeding Rate, Operating Duration and More

BRAINBOX XANDER

Warpage, Shrinkage & Displacement of Turbo Charger

Brainbox Xander helped to understand thermal, vibration and breakage points due to stress and deflection caused strains in the turbocharger. The system interprets and detects differential coefficients of expansions and alerts in advance 

​

• Transient Analysis in SI

• FEA Processes (thermal and mechanical)

• Brazed Joint Analysis with Stress result Visualization

• Elasto-Plastic Changes

• Knowhere has detected turbocharger fails and have predicted approx. time to fail

• The time to fail KPI has increased the uptime of the Hover Craft exponentially and have reduced maintenance cost by 70% overall

BRAINBOX KNOWHERE

Construction and Deconstruction of Stochastic Objects through Forensic Evaluation 

Detection of high-speed stochastic objects within minimal time from source location.

BRAINBOX KNOWHERE

Predicting System Failures

Knowhere used deep learning method to predict failures in systems using recurrent neural networks (LSTM). The system used novel feature representation of logs, during each time epoch in a low dimensional vector space. The system is configured to provide

​

• Predictive period

• Infected Period

• True Positive

• False Positives