The year 2023 will be the age of Edge Artificial Intelligence (AI). How and why?
Supply chain challenges, labour shortages and economic uncertainty had companies re-evaluating their budgets for new technology. For many organizations, Artificial Intelligence (AI) is viewed as the solution to a lot of the uncertainty bringing improved efficiency, differentiation, automation and reduced cost. Until now, AI has operated almost exclusively in the cloud. But increasingly diverse data streams are being generated around the clock from sensors at the edge.
These require real-time inference, which is leading more AI deployments to move to edge computing. The global Edge AI software market is set to grow from US$590 million in 2020 to US$1.83 trillion by 2026, according to a report published by MarketsandMarkets Research.
For airports, stores, hospitals and more, AI brings advanced efficiency, automation and even cost reduction, which is why edge AI adoption accelerated last year. In 2023, the industry expects a similarly challenging environment, which will drive the following edge AI trends.
Edge AI picks up momentum
A few years ago, AI was often viewed as experimental, but according to research from IBM, 35% of companies today report using AI in their business, and an additional 42% report they’re exploring AI. Edge AI use cases can help increase efficiency and reduce cost, making them a compelling place to focus new investments. For example, supermarkets and big box stores are investing heavily in AI at self-checkout machines to reduce loss from theft and human error. With solutions that can detect errors with 98% accuracy, companies can quickly see a return of investment in a matter of months.
IoT growth driving Edge AI
Per Statista, the number of Internet of Things (IoT) devices worldwide is forecast to almost triple from 9.7 billion in 2020 to more than 29 billion IoT devices in 2030. In 2030. However, it has always been a challenge to perform deep learning in low-power IoT devices due to the limited data storage and computational power of these resource-constrained devices. Now, Edge AI models are frugal enough to work at the Edge, empowering devices to complete their own data processing and generate insights without relying on cloud-based AI. Now that deep learning can be performed on-device, innovative new use cases are constantly popping up in many different industries.
Deep learning is a machine learning technique that trains computers to think the way humans do. Deep learning is accomplished when computers can learn by example. By taking advantage of multi-layered network architectures and huge sets of data to continuously learn, these models can reach new heights in terms of accuracy. There are times when they can even surpass what humans can achieve!
Deep learning trains machines to deal with any problem that necessitates the need to think, making it an essential technology across many industries. End devices such as IoT sensors generate large volumes of data that need to be analysed in real-time using deep learning. In turn, this data is also used to train deep learning models.
5G Edge AI is a must-have
When it comes to 5G, Edge AI is no longer ‘nice to have.’ Instead, it’s a must-have component due to its ability to handle the complexity of this 5th-generation mobile network. According to Precedence Research, the global 5G IoT market size is predicted to be worth around US$297.1 billion by 2030 and growing at a registered CAGR of 70.04% over the forecast period 2022 to 2030.
Edge AI (and new data processing and automation capabilities that come with it) support a diverse ecosystem of evolving networks in a way that cloud-based solutions cannot manage as effectively. Additionally, self-driving cars, virtual reality, and any use case requiring real-time alerts need Edge AI and 5G for the speedy processing it promises. As such, 5G is encouraging Edge AI adoption in numerous industries.
Human-computer interaction becomes an Edge AI use case
Often seen as a far-off use case of Edge AI, the use of intelligent machines and autonomous robots is on the rise. From automated distribution facilities to meet the demands of same-day deliveries to robots monitoring grocery stores for spills and stockouts to robot arms working alongside humans on a production line, these intelligent machines are becoming more common.
According to Gartner, the use of robotics and intelligent machines is expected to grow significantly by the decade’s end. “By 2030, 80% of humans will engage with smart robots daily, due to smart robot advancements in intelligence, social interactions and human augmentation capabilities, up from less than 10% today.”
For this future to happen, one area of focus that needs attention in 2023 is aiding human and machine collaboration. Automated processes benefit from the strength and repeatable actions of robots, leaving humans to perform specialized and dexterous tasks that are more suited to our skills. Expect organizations to invest more in this human-machine collaboration in 2023 to alleviate labour shortages and supply chain issues.
The Edge in Cyber Security
Cyber-attacks rose 50% in 2021 and haven’t slowed since, making this a top focus for IT organizations. Edge computing, particularly when combined with AI use cases, can increase cybersecurity risk for many organizations by creating a wider attack surface outside of the traditional data centre and its firewalls.
Edge AI in industries like manufacturing, energy, and transportation requires IT teams, to expand their security footprint into environments traditionally managed by operational technology teams. Operational technology teams typically focus on operational efficiency as their main metric, relying on air-gapped systems with no network connectivity to the outside world. Edge AI use cases will start to break down these restrictions, requiring IT to enable cloud connectivity while still maintaining strict security standards.
With billions of devices and sensors worldwide that will all be connected to the internet, IT organizations must both protect edge devices from direct attack and consider network and cloud security. In 2023, expect to see AI applied to cybersecurity. Log data generated from IoT networks can now be fed through intelligent security models that can flag suspicious behaviour and notify security teams to act.