Artificial intelligence (AI) will continue to drive innovation across industries in 2021, and AI at the edge is no exception. Indeed, ABI Research forecasts that within the next four years, the edge AI chipset market will reach $12.2 billion, surpassing the cloud AI chipset market.
In 2021, a new generation of high-performance, low-power edge AI processors will reach the market, making AI infrastructure increasingly accessible and affordable, while also offering lower latency and greater data privacy than the cloud. This shift toward the edge will provide organizations with AI solutions that are more powerful, versatile, responsive, and secure. Here’s my perspective on how different sectors will reap the benefits of AI at the edge in the coming year.
Smart cities: protection of data in motion
As the coronavirus pandemic has poignantly illustrated, the effort to keep the public safe and healthy increasingly requires governments to be tech- and data-savvy. To that end, investments in smart city technologies are on the rise, with Business Insider Intelligence projecting $153 billion in investments in 2021, compared to $112 billion in 2019. By 2025, that figure is slated to reach $295 billion.
A wide range of AI-enabled technologies fall under the smart city umbrella – such as tools for traffic management, apps that connect citizens to municipal services, cameras and sensors that help ensure public safety, and much more. Harnessing the benefits of these tools while maintaining robust privacy protection is critical – and to mitigate the risk that data in motion will be stolen or exposed, edge processing is crucial.
Smart cities’ edge AI infrastructure will also be essential to supporting expanded autonomous delivery services of the kind that different locales have experimented with during the pandemic. Because Covid-19 and the ensuing acceleration of e-commerce have increased consumer expectations for swift deliveries, demand for these services will continue to climb even after the pandemic is behind us, and better automation through AI-powered sensors – with data processed at the edge – will be vital to making this happen.
Industry 4.0 and smart retail: more efficiency at the edge
Efficient processing at the edge will also help usher in the next wave of innovation in the industrial sphere. Processing at the edge rather than in the cloud translates into significant cost reductions along with swifter, more efficient computing for tasks such as inspection, quality assurance, and better safety measures.
In 2021, factories will rely on cameras with AI edge processing technology to monitor equipment and production and, to the extent that social distancing measures are still required, to ensure that appropriate health protocols are being kept on the factory floor.
What’s more, utilizing AI at the edge will allow industrial robots – like the kind that Walmart is using to disinfect its stores’ floors and assist with inventory – to play a much greater role in day-to-day functions.
Retailers will also depend on edge AI solutions to operate more efficiently both during the remainder of the pandemic and beyond. Take cashier-less and checkout-free stores, for example. Edge processing solutions will help process video streams coming from multiple cameras on-premise in real time, without significant latency or high costs, making the customer experience with these new technologies that much more seamless, and enjoyable.
Remote healthcare: protecting patient privacy
Like other changes brought by the pandemic, telehealth and remote patient monitoring (RPM) were already set to play a greater role in future healthcare systems and economies, but Covid-19 dramatically accelerated that process.
Meeting soaring demand for healthcare and improving patient outcomes – to say nothing of protecting vulnerable patients from contagion – requires more use of RPM tools, and with demand for these devices set to double in the next five years, it’s crucial to ensure that the data generated by vital signs monitors, personal care robots, and other devices is both reliable and secure. This is particularly important for devices that assist the elderly, for example. If an elderly person has been injured at home, that information should be able to be sent quickly to the proper alert center, without the risk of any private information being exposed in transit. Edge processing will enable data to be processed faster, more reliably, and more securely – reducing the need for data to be transmitted to the cloud.
As enterprises across sectors look for solutions that ensure their devices are more powerful, versatile, responsive and secure, the cloud will continue giving way to the edge in 2021. Though a hybrid model will certainly take center stage for years to come, those who succeed in implementing AI at the edge will certainly gain an edge, quite literally, in a vast number of industries.
Orr Danon is CEO of Hailo Technologies. Prior to Hailo, he served in various hardware, software and system development project roles in the Israel Defense Forces. He has a Master of Science in electrical and electronic engineering from Tel Aviv University.
- Cartesiam IDE adds edge anomaly classification on Arm Cortex-M MCUs
- Edge computing is here: what’s next?
- Training AI models on the edge
- Ambarella targets intelligent edge sensing with new camera SoC
- Datafication and edge autonomy drive smart solutions
- Qualcomm preps edge AI offering
- Choosing solutions for edge AI
- Microcontrollers take on growing role in edge AI
- Edge AI platforms beat GPU/CPU on performance and efficiency
- Where’s the IoT edge in edge AI?