IoT edge AI is disrupting industrial markets

Artificial intelligence at the edge of the network is the cornerstone that affects the future direction of the technology industry. If artificial intelligence is the engine of change, semiconductors are the oil that will power the new era defined by machine learning, neural networks, 5G connectivity, and the emergence of blockchain, digital twins, and the metaverse.

Despite recent disruptions to the chip industry due to supply chain and macroeconomic factors, the convergence of artificial intelligence and the Internet of Things is poised to transform the world from cloud-centric intelligence to a more distributed one intelligent architecture.

It is expected that by 2025, the amount of data generated by IoT devices will reach a staggering 73.1 terabytes of data. As a result, endpoint data will grow at a compound annual growth rate of 85% from 2017 to 2025, driving intelligence from the cloud to the endpoint to run AI/ML workloads in tiny machines.

Some of the most disruptive applications include the development of “voice as user interface” to improve human-machine communication, as well as environmental awareness, predictive analytics and maintenance. Key growth areas include wearables, smart homes, smart cities and smart industrial automation.

What are the benefits of embedding intelligence in terminals? Many industrial IoT applications operate in environments constrained by memory capacity, limited compute and battery power, and sub-optimal connectivity. Additionally, these applications often require real-time responses, which may be mission and system critical. Expecting such devices and applications to run in a cloud-centric intelligent architecture is unfeasible.

This is the power of embedding intelligence in terminals, which is evolving from standard industrial IoT implementation to what we call AIoT for industrial applications.

Transforming data at the source of collection minimizes latency and enables optimized processing for time-critical applications. Since data is not processed and transmitted over the network, security issues related to data transmission and flow are greatly reduced.

Another advantage is that data processing can be connected to the endpoint's root of trust, making the implementation immune to attacks. Because data processing occurs at or very close to the source, we can take advantage of data gravity and reduce the power consumption associated with turning on the radio or moving data through the network.

Our commitment to our customers is to lead the industry in endpoint computing technology with the broadest range of MCUs and MPUs. This already enables designers to leverage our rich IoT ecosystem and AI/ML building blocks by leveraging the technology ecosystem, featuring more than 300 commercial-grade software building blocks from Renesas’ trusted partners.

Our growing AIoT portfolio also explains our recent acquisition of RealityAI, a new platform using Renesas processors to support edge and endpoint AI in industrial IoT applications.

Reality AI automatically searches a wide range of signal processing transformations and generates customized machine learning models while retaining traceability in its approach and providing valuable hardware design analysis. This model runs on nearly every MCU and MPU core Renesas offers, with new ones being added all the time.

This gives designers a very powerful tool to help them solve their most difficult problems, as the models are developed specifically for non-visual perception use cases and are based on advanced signal processing mathematics and edge deployment.

This enables advanced analytics to support a complete hardware design and a complete framework, including data collection, instrumentation, firmware and ML workflows. Other solutions simply generate algorithms and models, often accounting for only 5% of typical project costs, while ignoring the other 95% of development expenses.

Our comprehensive approach to AIoT design allows developers to reduce unplanned equipment downtime, increase productivity, and perform complex quality assurance tasks that are expensive or difficult to replicate in current test environments .

Automated ML Solution for Real-Time Analysis - AI for Advanced Signal Processing Endpoints

Tested on a 3-ton residential HVAC system under 51 different environmental and load conditions In real use cases, Realistic AI can achieve an accuracy of more than 95% when detecting and distinguishing single fault conditions. Testing also found OEM specifications for indoor and outdoor air flow obstruction and charging failure at 5% in both heating and cooling modes.

The integration of artificial intelligence and the Internet of Things in industrial applications is a major trend with huge potential. The acquisition of Reality AI unlocks the potential of combining advanced signal processing with AI, supported by Renesas’ rich hardware, software, tools and ecosystem, providing all the building blocks needed to unleash creativity.

The above is the detailed content of IoT edge AI is disrupting industrial markets. For more information, please follow other related articles on the PHP Chinese website!