Motion intelligence technology
Motion intelligence is about understanding how things and people move. This can be done using a range of different technologies.

We translate raw sensor data from motion sensors (accelerometer, gyroscope, magnetometer) on small electronic devices into useful information that it is possible to act on, by applying our own artificial intelligence technology on the data. This is deep tech.

We have invented a unique solution, which means that our motion intelligence software runs on small devices. This is a different approach from solutions where all sensor data is streamed from the device to the cloud, and where all the intelligence is in the cloud.

Advantages

Recognise motion in real-time with high accuracy

An important advantage is that we can recognise motion in real time with high accuracy on the device, which means that we can immediately act on certain situations or deviations from normal behavior. 

Intelligent and autonomous

The device is intelligent and autonomous, which means that it can think for itself and it does not have to be connected to a network all the time. It can be disconnected and connect when a certain movement or situation occurrs.

Less data to send

Having the intelligence on the device lowers the amount of sent data. This lowers power consumption, because running the software on the device consumes much less power than transmitting all sensor data over Bluetooth LE (BLE) or any other communication method. This will improve battery time and it will lower the costs for data transmission, which is important in most IoT applications.

And the good news is that we can do it on low-cost processors, such as ARM Cortex M0. Cost is very important when you want to roll out millions of devices.

Easy to learn

SensorBeat learning is easy and fast and can be done with small datasets. Many other AI systems are slow to learn and require large datasets.

Small footprint

SensorBeat Lib requires around 10 kB of RAM on the device.

SensorBeat and LPWAN's

SensorBeat is a software product and is independent of hardware and type of communication. Using SensorBeat together with Low Power Wide Area Networks (LPWAN) would be an ideal fit, since they share the same qualities that are important for IoT networks, low power and low cost devices.

 

 

How does it work?

The first step in a SensorBeat project is to capture sensor data, "Capture". We use a smartphone with a mobile app and a capture device comprising sensors and BLE to capture sensor data and video. 

If we want to detect the motion of a chainsaw, we mount a capture device on the chainsaw. The capture device is streaming synchronised sensor data and video over BLE to the smartphone. Once data capture is ready, the recorded data is transmitted from the smartphone to a server for data analysis.

In the second step, "Learn", we use the recorded data as input into SensorBeat Analyser. Analyser is like a video recorder for sensor data and video. The sensor data is processed by SensorBeat to create patterns. Data, patterns and video can be rewinded back and forth in order to find the patterns for different movements. Once the data analyst finds a pattern for e.g “kickback” this is marked and stored in the system. Now, SensorBeat has learned the movement “kickback” and has created the "kickback" pattern.

In the next step, SensorBeat Lib and the patterns are installed on the hardware devices that will be shipped for commercial use.

Finally, when the hardware device is mounted on the chainsaw and the chainsaw is in operation, "Use", it is constantly listening for the "kickback" pattern and other patterns that were generated during the learning phase.

 

SensorBeat

SensorBeat is a system

SensorBeat offers motion intelligence on the edge. This means that we can offer high accuracy motion recognition in real-time, on low-cost hardware and with low power consumption.  

The key advantage of motion intelligence on the edge of the network is to leverage the benefits of analyzing real-time motion data, without the bandwidth costs that come with sending that data to the smartphone or to the cloud for analysis. This has a clear value from a power consumption perspective, and means that we can offer a longer battery life. Battery life is critical in many IoT applications. 

SensorBeat can feel with great accuracy, and that makes all the difference. Higher accuracy means for instance that SensorBeat can detect not only that you are walking, but also estimate your speed or the fact that you are going uphill or downstairs.

SensorBeat is protected by granted patent, publication number 578 370. SensorBeat is a registered EU Trade Mark with registration number 014868426.

SensorBeat Analyser

SensorBeat Analyser is a software package for analyzing and collecting action from motion sensors and video.  Analyser can playback video with synchronised sensor data. Analyser is where you teach the system to understand specific movements, and it will create the pattern database for the embedded version of SensorBeat. 

Analyser can be operated by Imagimob or by our customers.

SensorBeat Capture

SensorBeat Capture comprises a mobile app and a capture device. It’s used to capture synchronised sensor data and video during the data capture phase. Once the data capture is ready, it is sent to SensorBeat Analyser for analysis.

SensorBeat Lib

SensorBeat Lib is a software package that can be easily integrated into 3rd party hardware platforms. SensorBeat Lib is fast, accurate and has a very small footprint, which means that it can run on low-cost and low-power hardware devices. For OEM devices SensorBeat Lib is a source code / pre-compiled library written in ANSI C.

SensorBeat Lib is also available on iOS, Android and Apple watchOS.

SensorBeat Lib is communication-agnostic, ie it works with all communication technologies in the market, eg Bluetooth LE, 2G/3G/4G, WiFi, Sigfox, NB-IoT and LoRa.

SensorBeat Cloud

The SensorBeat Cloud is used to download OTA updates of SensorBeat Lib and pattern databases to devices. It also allows for storage of usage data from devices. 

SensorBeat Device

The SensorBeat Device is a complete device that contains all the necessary electronics, firmware and artificial intelligence software (SensorBeat Lib) for understanding the motion of things and people. The device is programmable, it can easily learn new motions.

It's an optimal fusion of hardware and software. It contains top of the line sensors, impressive storage capabilities, Bluetooth LE, a rechargeable battery and external communication ports. 

The SensorBeat Device is intelligent and autonomous.

Using the SensorBeat Device will remove the risk of firmware and application development for customers, and therefore lower costs and shorten time-to-market.