Survey: AI takes on a bigger role in IoT development

(Source: Newark)

Engineers are more willing to deploy artificial intelligence (AI) to improve their products and to increase the return-on-investment of their systems, according to Newark’s (an Avnet company) fourth annual Global IoT Survey. The survey of 2,263 engineers and others involved in the development of IoT solutions also finds that companies are establishing partnerships to drive the development of new IoT solutions.

The survey reveals significant growth in the use of AI over the past year. Survey respondents using AI increased to 56% from 39% in the previous year’s survey, with 26% expected to consider using AI in the future. The most popular use for AI is preventive maintenance (16%) followed by robotics (12%), image classification (11%), and business analytics (10%).

Engineers are also finding more applications for AI in their IoT designs. Respondents who thought that AI was not useful dropped from 32% in 2021 to 13% in 2022. Only 3% said they didn’t understand AI compared to 13% last year.

Here are some key findings that may be useful for both IoT developers and system integrators.

The survey finds that companies are discovering new and diverse ways to adopt IoT across the different industry sectors. Although 30% of survey respondents cited productivity and manufacturing improvement as the primary value of IoT, there are a variety of IoT applications being used, including business intelligence & data analytics (18%), customer care (17%), operations (15%), and mobile apps (11%).

Survey respondents believe the top IoT industries in five years will be Industry 4.0 (18%), followed by energy management (16%), home automation (15%), AI (14%), and automotive/transportation (11%). When compared to the previous year’s survey, energy management is making some headway, increasing from 10%.

Sensors play a big role in IoT development and are used across a wide range of applications. The most popular application by far is environmental sensing (34%), following by motion sensing (15%), opto/image sensors (15%), location sensors (10%), health monitoring (8%), and audio detection/recognition (5%). However, 13% of respondents reported they do not use sensor technology in their IoT designs.

 When asked about their key concern regarding IoT implementation, security (37%) is their biggest concern, followed by connectivity (29%), interoperability (22%), and ecosystem (10%).

In terms of hardware platforms, survey respondents overwhelmingly use single-board computers or SBCs (50%). This is followed by custom hardware (31%) and manufacturers’ development boards (19%). Their most preferred SBC is Raspberry Pi (46%) followed by Arduino (26%). When it comes to development platform vendors, STMicroelectronics (28%) and Microchip (23%) topped the list. Other vendors cited include NXP, Avnet, Infineon, Xilinx, and Renesas.

Preferred development platform (Source: Newark)

The survey also reveals that there is a growing trend towards working with partners, which are critical for most companies developing IoT solutions, said Newark. Findings show that 34% of companies never take the lead role in development, matched by sometimes taking the lead role, indicating that there are strong relationships between organizations, said the company.