The design of the sensor characteristics is supported by the Confusion Matrix, which has its origins in machine learning models. The Confusion Matrix, also known as the error matrix, is used to clearly describe the detection behaviour and limits of sensors. The illustration shows the general representation of a confusion matrix (own illustration).
Sensor development using Confusion Matrix
As a provider of complete sensor solutions, STEINEL Solutions has many years of experience in the development and production of sensors in a wide range of industries. In addition to experience, successful implementation is also based on the Confusion Matrix methodology. A technical article on sensor development using Confusion Matrix.
Definition of the use case
The detection behaviour of a sensor application is crucial to the success of the product. The different application situations, known as use cases, define the behaviour of the sensor and describe how the sensor should react in each case. Not only the ideal case is considered, but also all environmental factors.
STEINEL Solutions AG has brought successful sensor solutions to the market in many sectors. As a result, various application situations and solutions, such as in the area of people detection, are already known and implemented. If a new application deviates from existing solutions or new use cases are added, a joint development process begins between the product owner and the realisation partner.
- The product owner knows the industry-specific application and environmental conditions of their product. He specifies the requirements and the desired behaviour, can also describe unfavourable installation situations thanks to his experience, interprets test results of the first prototype and derives measures from them.
- The realisation partner develops the optimum solution with its broad-based expertise. This is usually an embedded system with sensors and actuators or any communication interfaces.
An iterative process in the intensive validation of early functional samples and more mature prototypes brings sensor optimisations and product improvements to light. In some cases, use cases are even redefined during this phase. Before the project moves on to industrialisation, the requirements are jointly coordinated with the test descriptions using a test plan and then verified point by point.
Configuration of the sensor characteristics using Confusion Matrix
The following Confusion Matrix describes a passive infrared (PIR) sensor using the example of a motion sensor as used in outdoor areas of buildings, among other things. These are not real images, but illustrative examples for understanding the matrix. It is often advisable to define the frequency of occurrence of the individual use cases as a percentage with the product owner and set an acceptable value, e.g. False NEGATIVE is tolerable up to 0.5%.
True POSITIVE: These cases are usually defined quickly and reflect the actual function of the sensor. They are the situations that the sensor should recognise. A motion sensor recognises people walking past at a speed of >2km/h and switches on the light.
True NEGATIVE: Seems pretty obvious at first glance. Is it the opposite of True POSITIVE? Simply put, the motion detector should not detect a person in its detection area who is not in the room. But what about a seated person who is not moving? In this case, a motion detector is actually referred to as True NEGATIVE. In practice, a distinction is made between presence sensors (typically used in offices) and motion sensors. This explains why the motion sensor, unlike the presence sensor, does not react to a seated person. However, it has a greater range and cheaper components.
There are also two fields: False POSITIVE (phantom detection) and False NEGATIVE (blindness). Both cases describe a malfunction of the sensor. The product owner and the development partner must agree on the degree to which such malfunctions are permissible. This delimitation is very individual and can have an impact on development and product costs. For example, additional sensor technology may be required to reliably recognise specific special cases. Whether the application accepts the additional cost of the additional principle ultimately depends on the product owner's price-performance assessment.
Interaction of many specialist areas in the realisation
Many specialist areas come together in the technical implementation of a sensor application: analogue design, digital filtering and signal processing, EMC immunity, power management, communication connections, adjustment and calibration, etc. All these areas require the appropriate expertise, infrastructure and tools to adapt existing technology modules to customised solutions.
STEINEL has specialised in the field of PIR, high-frequency and optical sensors and has become the market leader. The experience and resources in the areas of hardware and firmware development, PCB design, mechanical development, application testing and manufacturing and process expertise are correspondingly extensive.
A sensor usually contains an analogue circuit. This makes the PCB a customised component. The PCB designer works closely with the hardware developer to develop the optimum layout. The sensor must also be protected from environmental influences and has requirements for its assembly - depending on the technology or as defined by the product owner. These aspects are incorporated into a mechanical design, which in turn places demands on production processes. Here, the interaction between design and production is a central process that should not be underestimated, as it eliminates critical interfaces in a solution that is developed and produced in-house.
The picture shows Andreas Münger in the EMC laboratory.
The firmware is a very central part of a sensor. Analogue sensor signals are digitally converted and filtered and processed in a microcontroller. In autonomous sensors, the firmware processes the information from the sensor and controls an actuator, for example a valve for a washbasin tap.
In a networked application, the sensor information is made available to a higher-level system via an interface. As a result, spatially distributed sensors can enable much more complex detection behaviour. With the wide range of communication technologies, sensors have developed strongly towards the IoT (Internet of Things) in recent years. A wide range of implemented solutions allows STEINEL to select the right technology for a variety of use cases.
Sensors that are operated with a battery should function autonomously over a long period of time. Low-power applications are the rule today. Battery management can be just as much a part of this as sophisticated power-saving tricks.
Another discipline that should not be underestimated is the standard-compliant implementation and certification of the product. EMC immunity is always a critical area for a sensor. STEINEL Solutions uses internal pre-compliance tests in its own EMC laboratory to optimise false positives and false negatives. The solution often lies in clever signal processing in the firmware, which ultimately also results in cost neutrality for the product. Close coordination between hardware and firmware development is crucial for an optimised overall solution.
‘What fascinates me as a developer of sensors is the extreme breadth of specialisms that are required. Many years of experience enable me to implement hardware and firmware myself and to find optimal solutions in both areas. A good understanding of the application in the market is a prerequisite for this.’