Industrial and commercial Machine vision is rapidly becoming a key enabling technology for digitalization and automation in automotive, healthcare, manufacturing, retail, smart buildings, smart cities, transportation, and logistics.
According to ABI Research, the total revenue of machine vision technology in the seven major markets is expected to reach US$36 billion by 2027, up from US$21.4 billion in 2022. This growth translates to a CAGR of 11%.
Traditionally, machine vision was primarily focused on surveillance and security, asset monitoring, and defect inspection. These mature applications continue to drive the main bulk of total camera shipments within the enterprise market. However, the industry is going through an exciting phase. The Covid-19 pandemic and the desire for digitalization have led to the emergence of innovative use cases, such as occupancy detection, crowd monitoring, predictive maintenance, high precision automated inspection, automated picking, and sorting systems in warehouses.
“These innovative use cases are expected to drive future growth of the industry. A key enabler of these innovation use cases is Machine Learning (ML), particularly Deep Learning (DL) technology in machine vision. Most, if not all, technology suppliers offer DL-based solutions that are flexible, scalable, and highly efficient. At the same time, enterprises are waking up to the benefits of DL-based machine vision,” said Lian Jye Su, a Principal Analyst at ABI Research. “When combined with factors such as decreasing component and engineering cost, increasing ease of integration with third-party solutions, growing open-source software and toolkits, the barrier to adopt an effective machine vision solution has lowered significantly for many enterprises.”
Moving forward, distributed computing will become the central theme in the implementation of ML in machine vision. Computing platform vendors, such as Intel, NVIDIA, Qualcomm, Xilinx, and NXP, have been actively launching processors that can run ML models on cameras directly, or on gateways and on-premise servers that cameras connect to. Instead of having ML models running in the cloud, these vendors have developed a suite of solutions ranging from ML processors to ML development environment and embedded security enhancement to ensure timely development and deployment of ML models and smooth integration into existing workflows. Furthermore, this domain is expected to become more competitive with the emergence of innovative startups focusing on machine vision at the edge, such as Hailo, Perceive, Syntiant, Mythic, GrAI Matter Labs, and DeGirum.
Currently, hardware revenue is the main component of the revenue at around 89%. However, the share of software and services is expected to grow over time, growing from 11% to 16%. With the emergence of DL-based machine vision, more and more ML solution providers are likely going to build their revenue models around the development, deployment, and maintenance of vertical-specific DL-based machine vision models. For example, Instrumental and Landing AI focus on machine vision solutions for manufacturing, Cipia and Cogniac in video telematics and driver monitoring, Arterys and Lunit in patient diagnosis. “Instead of fully relying on internal expertise, enterprises can partner with these companies to develop targeted solutions together. Such partnerships are critical in reducing the complexity in building and maintaining custom ML models, accelerating time-to-market while maximizing Return On Investment (ROI),” concluded Su.