automates analysis of complex image or time-series data to provide continuous, objective measurement of how changes in manufacturing affect product performance.
R&D focused & scalable - goal: customize quickly, but provide an end to end solution
experience with battery R&D - goal: understand battery failure modes
integration of telemetry & image data - synthesize images into quantitative data
robust algorithms that are resistant to sensor failures
We are building a software product to use cloud-based machine learning to better understand telemetry & image data from complex systems, getting products to market faster and reducing costs.
Ultimately, Machinery Analytics intends to leverage IoT and Industry 4.0 to launch a transformation in manufacturing quality assurance and control from the current method of ensuring process repeatability to a new paradigm where product performance is the driving force for process and design decisions.
Our product is designed to be massively scalable, so that the insights gained during R&D can be used to monitor and improve performance during full scale production, or ultimately used to provide failure identification and prediction for end users of high volume, high value products.
INTERESTED IN OUR PLATFORM?
Let your engineers and scientists focus on what they do best - bringing their visions to life
Next Generation Manufacturing Canada
FREQUENTLY ASKED QUESTIONS
How do I connect my data?
We provide a secure link to storage hosted by google-cloud. You can save .csv files directly to this location. Later implementation will allow direct connection to distributed data acquisition systems eg. vehicles or remote sites.
How secure is my data?
Why do I need Machine Learning?
Machine learning looks at all of your data to create a mathematical model, or fingerprint of your device. That model is used to distinguish real changes in performance from environmental, load or user inputs.
What if I want to develop my own quality control AI/ML tools?
We give you the ability to export a TensorFlow pipeline from your data set, including batched input data files, as well as prototypes to add your own engineered features, train and evaluate.