Enabling Collaboration to Unlock IoT's Boundless Potential
The survey reaffirms that everyone from SoC designers to OEMs, engineers to software developers, vertical sectors across the board sees enormous opportunity ahead for the IoT. And it illuminates clear pathways for scalability. These include:
Reduced software complexity.
Development and security standards to drive compatibility.
The need to leverage new technologies such as 5G, the metaverse, machine learning, cloud-native development and edge computing.
Industry collaboration to drive greater compute performance with security by design.
Standards and software and libraries to reduce fragmentation and other software challenges.
Improving the developer experience to reduce time to market.
Follow the moneyThe top three IoT sectors expected to generate the highest levels of revenue over the next five years are consumer technology, the data center/cloud and then smart cities. Network infrastructure and industrial sectors weren't far behind. In any case, three key sensor technologies underpin all of these sectors: Vision, voice and vibration. (Think cameras, smart assistants in the home and on smart phones and systems that monitor industrial equipment). Of these, vision is the most technically complex but also draws the most interest across all sectors. Vision-enabled platforms will be everywhere improving safety, security and productivity.
Arm's Paul Williamson on the future of IoT at Embedded World 2023
When asked about the technologies or product categories that will have the most impact on the IoT over the next five years, 5G was the top choice, with 46% of respondents selecting it. This was followed by the metaverse at 41%, cloud-native development at 38%, machine learning at 38%, and edge computing at 29%. While machine learning ranks third, an analysis of the data suggests its potential looms much larger for respondents.
Learn how smart vision devices accelerate development for the entire value chain.
For example, for the question around what are the ways in which a company's time-to-market might be accelerated, 33 percent answered more vision-based computing resources and development support, which are inherently AI-powered.
Another reason AI looms large in IoT development is that respondents want to improve the use of data analytics in their business decision making. We're seeing ML-enabled IoT systems deployed in commercial buildings to improve foot-traffic flows and conference room usage, not to mention improve HVAC efficiency; parking structures are removing scores of cameras that identify open space and replacing them with two, one at the entrance and one at the exit to analyze cars and infer open spaces.
The most important reason for developing IoT systems leveraging AI is to improve the customer experience, and leading-edge retailers are beginning to use IoT systems to analyze consumer behavior inside their stores to optimize layout, product presentation and the checkout experience.
The advent decades ago of the system-on-chip (SoC) methodology revolutionized system design. It put into place a modular approach to creating bespoke silicon that could enable innovators to focus on their expertise and leverage the design ecosystem for other elements. The importance of the model hasn't dimmed with time. A significant majority (nearly 90%, below) of respondents consider SoC integration to be important in their design process. SoC integration, which, over the years, has been propelled forward by the Arm architecture, has revolutionized how companies get their ideas to market quickly and efficiently.
Secondly, an overwhelming 90% of the participants (above) emphasize the value of basing their products on the same hardware architecture. In an increasingly complex design and development environment with relentless market pressures this shouldn't be too surprising. Targeting the same hardware gives development more scalability, ensures
trusted suppliers, and reduces costs and learning time.
It speeds time to market because developers can re-use the knowledge gained on earlier designs to build new ones on the same architecture, enables software re-use and, in some cases, is a customer requirement.
