What used to get dusty in folders is now saved in form of zeros and ones on hard disks. Digitalisation is leading to an ever-increasing data collectomania. However, the mere existence and collection of data does not automatically create value.
As Internet of Things provider Microtronics has been connecting products, machines and services with the Internet for 14 years. In more than 300 successfully realised projects in more than 80 countries, data is recorded, transferred and stored together with international customers. With the help of this information for example people are warned of flooding, delivery routes are optimised or free parking spaces are displayed. Microtronics is fulfilling its role as a technology leader with an R&D quota of 20%.
The potential of Edge AI
Microtronics is currently working on AI (Artificial Intelligence). AI describes an automated, intelligent and self-learning behaviour of machines. Such applications require a lot of computing power and energy.
“On the edge”, i.e. directly on the device (e.g. a fitness tracker or an energy meter), a low amount of energy is available. With a small battery, the devices should run for a long time. This is why the logic for AI applications is usually executed on the server or on the IoT platform. There computing power, energy and memory space play a subordinate role. For this, the large amount of data must first be transferred to the IoT platform. Energy is required for each transmission. In the case of data transmission via mobile radio costs are also incurred.
Microtronics is currently working on Edge AI – i.e. processing the data directly on the device using artificial intelligence and reacting to the result accordingly. In the consergio project, keyword spotting (the speech recognition of defined commands) takes place directly on the device. It is not necessary to transmit the entire audio recording to the IoT platform, but only the recognized command. This is an enormous advantage in terms of security and data protection.
Older modules of electricity and water meters often have no interfaces for digital recording. These are retrofitted with Edge AI. The meter reading is photographed. The algorithm recognizes the photographed numbers and only transfers the numerical values instead of the photo.
With regard to latency, Edge AI is interesting for various applications. A striking example is an autonomously driving car. Even a few seconds delay can decide between life and death. It would simply take too long to first transfer the recorded data to a server, analyse it there and then transfer the result and the resulting command back to the car. Edge AI is therefore indispensable for autonomous cars.
Expansion of technology leadership
In order to continue to successfully realise IoT projects with customers, Microtronics will continue to invest in research and development in 2020. AI will play an important role, both on the end device and the IoT platform.
Due to the new possibilities, it is still necessary to design the application exactly before the technical realisation. When, where and in what granularity the data is needed is only an excerpt of the necessary considerations.
With the Microtronics IoT Suite you get a comprehensive toolbox with all modules for the realisation of your own IoT applications. Modular you choose what you need and benefit from the regular expansion and integration of state-of-the-art technologies.
- 8. IoT Forum – Let’s go hybrid - 09/09/2020
- How IoT supports emergency services and citizens during floods - 07/14/2020
- Why AI supports human intelligence but does not replace it? - 06/25/2020