Transformative insights using sensors and AI machine learning
View the 60 second video to learn about the advantages of using machine learning with sensor data.
Identify situations that were previously thought to be difficult or impossible to detect. Classify, predict and react for a step change in efficiency and safety.
Sensors allow a large number of physical factors to be quantified and analysed.
AI machine learning excels in situations where the data is too complex to be analysed by a human or conventional computing algorithm.
We supply software that runs on edge devices, within range of sensor beacons, detecting location, movement (accelerometer), proximity, temperature, humidity, air pressure, light, magnetism, open/closed, proximity, heart rate, fall detection, smoke, gas, water leak or custom sensor signals.
The outputs are alerts sent via, for example, email and http.
Serverless Edge Devices solve many of the problems with Internet of Things (IoT) cloud solutions:
Removes the need for cloud storage of redundant data, saving cost and resources
Solves the problems of prohibitively large data transmission required of higher sensor sampling frequencies
Brings computing much closer to, for example, patients, machines, customers, suppliers, employees, vehicles and buildings for quicker notifications
Secures your data as it doesn't go through, nor is held by, a third party nor is dependent on the reliability, availability and variable cost of a 3rd party service
We use AI unsupervised machine learning to extract features, for example, common shapes and patterns, in complex sequential sensor data.
The example on the left shows changing numbered features found in sequential incoming data that we use for detection, classification, anomaly detection and prediction.
As a simple example, 71 and 82 being active might signify an asset is being under-used, a machine has unusual vibration or predict a vulnerable person is about to fall.
Discover if we are the right people to help build your AI capability
Identification of your problems, the requirements, the required accuracy, performance and machine learning output actions
Implementation of sensing, data capture, cleaning and shaping
Development and refinement of machine learning model
Implementation of actions based on the output of the machine learning model