AMP’s Core Technology — AMP Robotics
AMP’s Artificial Intelligence:
A Neural Network that’s Unmatched
Millions of parameters, superior accuracy
AMP’s proprietary AI technology is unmatched in the industry, using millions of parameters compared with a fraction of that for competing systems. AMP’s breakthrough AI platform, AMP Neuron
What is artificial intelligence?
Artificial intelligence (AI) is a collection of algorithms that enable a computer to learn from experience. It’s an approach to teaching machines to do what humans can do. Humans are adept at identifying objects from visual inputs, whether it’s an image or views of the real world, and it’s this ability that enables us to perform complex tasks—like sorting materials in a recycling facility.
Machine learning is a subset of AI; AMP’s technology is based on a type of machine learning known as deep learning. Neural networks are deep learning technologies where a system learns how to build patterns out of other patterns and perform a task by analyzing examples.
How does AMP apply AI?
AMP’s proprietary AI technology works by perceiving images of conveyor belts within recycling facilities. Looking for specific colors, shapes, textures, logos, and more to recognize patterns correlated with material type, our AI platform learns to identify objects in the same way a human does. It’s this ability to construct a “model” of an object using simple building blocks like material features that makes deep learning so powerful.
Our team of several dozen data annotators examines these images and categorizes the objects belonging to each material type through a proprietary tool AMP created to offer the most efficient annotation process in the industry. We then train our neural network by showing it examples of items belonging to each material category, each new image at our growing number of installations contributing to the strength and effectiveness of our machine learning algorithms within an ever-expanding dataset. A robot installation in a California sorting facility is helping robots in Colorado, Minnesota, Florida, or any other location become better at identifying items that can be recycled.
What’s AMP AI edge?
Behind our world-class technology is a team of machine learning experts with deep backgrounds in software development, and a passion for continuous improvement.
We don’t outsource AI development to third parties; it’s the core technology of our business that delivers industry-leading precision and effectiveness. Our short cycle between research and development and deployment means our customers benefit quickly from our enhancements.
Our system consistently achieves an accuracy rate of 99% on the material it positively sorts, outpacing the performance of competing technologies like optical sorters and those using near infrared recognition or hyperspectral cameras. With superior accuracy, Neuron is capable of recognizing the brand of packaging or even the SKU (stock keeping unit) on material targeted.
Our neural network operates with millions of parameters, compared with a fraction of that for competing systems. The algorithms behind our AI incorporate millions of image examples for training, along with some of the strongest computing power available. This translates to a more expansive “capacity to learn” for our AI platform—and provides the foundation for our unmatched neural network.