Infera AI: Platform for Digital Twins and Better Decision Making

A software platform that leverages statistical inference and machine learning to make more informed decisions.

Through a novel combination of deep Bayesian learning and Bayesian analytics, the platform delivers accurate prediction statistics for robust, quantifiable, and interpretable analytics to scenarios that were previously too expensive or too slow.

Product features:

  • Supervised learning algorithms for high dimensional datasets
  • Predictive data-driven modelling with Bayesian confidence intervals
  • Efficient exploration strategies for identifying promising designs
  • Uncertainty quantification
  • Visualization tools for high-dimensional input-output relationships

 The platform can be applied to a range of sectors where impacts could include:

  • Providing a means to quantify risk, enabling alternative options to be rigorously traded
  • Identifying robust “optimal” decisions
  • Discovering relationships between decision variables and system performance metrics
  • Certifying models for deployment in high-stakes situations
  • Answering questions such as “do we have enough information to make a decision?” or “what should we do to become more informed?”
  • Decision making under uncertainty
  • Fault diagnosis and predictive maintenance

OPPORTUNITY

Current platforms and existing products have the following shortcomings:

  • Inability to deal with high-dimensional decision landscapes 
  • Expensive computer simulations do not scale to large and diverse data sets
  • Inability to account for uncertainty
  • Lack of error bars for decision support and explainability

Applications:

The software is especially applicable to digital twins in engineering, and can be used in the design, certification, and monitoring of assets.

  • Engineering design, optimization, risk management, and monitoring for digital twins including in aerospace, automotive, and materials
  • Oil & mining (using geostatistics for prospecting)
  • Healthcare (biomedical design and drug discovery)
  • Financial modelling and planning
  • Utility and resource management

STATUS

  • A startup is being formed to commercialize these tools with initial customers
  • Related production-level tools developed by the inventors are being used by leading aerospace companies for applications in engineering design

ID:

p2174

Keywords:

Software , Deep Learning , Data Analytics , Aerospace , Computer Graphics , Machine Learning , Digital Twin , Companies

VPRI Contact

Donna Shukaris

Innovations & Entrepreneurship Manager
Innovations & Partnerships Office (IPO)
(416) 946-7247

Related Resources