Three Areas of Competence
Optical Sensors
Optical sensors are powerful analytic tools as they are capable of providing analyte information quickly and remotely. Their basic mechanism is that photonic molecules interact with the target analytes under study and provide information about parameters such as the structure of the analytes. When the analytes molecules enter the system, the sensors produce detectable changes in their signals – which are then transduced into easily measured and quantified optical signals. Compared to other types of sensors such as electrochemical ones, optical sensors are particularly useful for non-invasive detection, which is critical for certain use cases. It is these use cases that we approach together with our customers and research partners.
Machine Learning
Surely you already have an idea of this technological approach, which is shaping our time like no other. And if not, let us recap: Machine learning (ML) is the idea that generalized algorithms can tell you something important about a collection of data without having to write any particular custom code to the problem. Instead of writing code, you feed the generic algorithm with data, and it constructs its own data-based logic. Our approach is to use the latest ML techniques and associated know-how to efficiently evaluate information that we obtain using optical sensors. Our ML, which is developed in the latest relevant programming languages, goes far beyond conventional applications. It distinguishes highly difficult signals and assigns them to correspondingly difficult patterns. This level of definition is the critical added value for our clients and partners.
Actionable Coupling
Of course, the two areas must not stand unconnected next to each other. In order to create really useful solutions for real use cases, an integration of hardware (optical sensors) and software (ML) must be achieved. It involves the integration of specific hardware devices including design and manufacturing of all supporting systems for the optical sensing device. This is coupled with integration with edge computational devices where the the ML algorithms are are invoked to provide immediate classification of the sensor output. Sensor systems are integrated with fixed and mobile communication systems utilising combinations depending on need of of Bluetooth, Bluetooth low energy Wi-Fi , 3g, 4g, 5gn NbIoT and LoRA. It is precisely these intermediate areas that decide whether applications remain in the laboratory or make the leap into practice.
