Many companies are trying to integrate AI into applications. However, many are still at the very beginning of product development. AITAD demonstrates how companies can take advantage of the integration of artificial intelligence into electronics when designing products.
Embedded AI refers to electronic systems in which artificial intelligence (AI) operates autonomously and locally. The market potential is quite large – in part due to follower trends such as (I) IoT, corresponding connectivity, security and cloud services. Allied Analytics estimates that the AI semiconductor market will be worth more than $190 billion in 2030. For comparison, the growth of the AI-as-a-service (cloud) market is estimated to be around $44 billion over the same period. Embedded AI is still only at the beginning of its development potential, at the current stage giving each product a unique selling point (USP). However, it is always important to The use and benefit of the manufacturer and the user in harmony with each other be.
Practically speaking, embedded AI can be divided into three main application groups:
- Functional innovations
- user interaction and
- Predictive/preventive maintenance.
The first enables new functionality that improves or even changes the intended utility of a product or process. As an additional output field, user interaction getting help from external sources. It ranges from simple voice command entry (such as KWS, keyword detection) through gesture recognition to complex human-machine collaboration such as operator tracking, eye tracking, or work piece detection. Typical maintenance topics such as predictive maintenance or preventive maintenance, which go beyond simple condition monitoring and provide early, intelligent predictions about specific fault patterns, are currently seen as probably the greatest “hidden needs” of many product manufacturers.
“Most companies often do not know what capabilities are in their products,” explains Vyacheslav Gromov, general manager of embedded artificial intelligence provider AITAD. “We have a lab where we use modern technology to collect data, but also produce the devices in a few hours and test them for serial production. Manal A variety of ways to integrate embedded AIMost of our customers are clearly surprised. However, only the components that have the greatest possible benefit to the customer and user are included in the customer prototypes. The market for embedded AI is still very murky in some places. There is definitely a need for more clarification here – from management to developers.”
Embedded AI has huge potential
Cloud AI alone is just a transition, and Gromov is sure the future lies in decentralized processing: “We’re working on the sensor on the circuit board with massive amounts of data that we haven’t been able to transfer further. The AI must process it and dispose of it directly on sitein order to track down the desired deep links.
Built-in AI makes it possible to process large amounts of data locally, reducing the risk of sensitive data being intercepted or tampered with. This leads to higher data and system security. The device does not have to provide a high-performance network infrastructure to be able to process data. Thus, it requires less contact, which reduces production costs. Embedded AI lives on limited resources, in terms of power supply (including battery operation), computing and storage capacity. These components collect and process data instantly and can interact with it in milliseconds, which is a must in many applications. The device can also analyze data in real time and transmit only what is relevant for further analysis in the cloud (keyword: reduce data volumes).
Tailor-made built-in AI is key
The embedded AI market is still largely unoccupied, with more and more isolated solutions or low-threshold offerings being added. Specific solutions (often also closed source) can be an advantage for the company in individual cases and if integration occurs at an early stage. Low threshold software offerings from various semiconductor manufacturers or more comprehensive tools such as “Edge Impulse” or “NanoEdge AI” are both a blessing and a curse: You will get the result quickly (Also in part thanks to the AutoML function, that is, the automated model generation process), the entire development chain, which depends on the understanding of the developer involved, is limited.
The semiconductor industry also offers a number of chips designed for embedded AI use cases such as image processing, which are both performance and adaptive. AITAD also participates in some research programs and promotes this disruptive trend in basic research. But even this diversity in the hardware market brings with it confusion.
»We clearly recommend that companies take an approach away from off-the-shelf solutions. It can only be adapted to needs to a limited extent, with smaller or larger cuts. Individual system productions have a much larger scope. This means knowing which AI model fits the product, how it can be effectively implemented on hardware, developing the corresponding system components based on the data collected and evaluated, implementing everything with a prototype and testing it in practice. This seems like a lot of effort at first. But if you look at how long the product has been on the market and the benefits companies and users have of it, for example in the area of preventative/predictive maintenance, the investment is definitely worth it,” Gromov continues.
Artificial intelligence market size, share | Analysis-2030 (alliedmarketresearch.com), accessed 5/5/2022 at 4:00 pm.
Artificial intelligence as a service market trends and growth forecast 2030 | MRFR (marketresearchfuture.com), accessed 05/05/2022 at 16:00.
You may also be interested in
Two 22 . Compact Computing