Predictions For A Year Of Pivotal Growth For Computer Vision
Carlos Anchia, Co-Founder and CEO at Plainsight
As we conclude 2021, the arrival of Omicron suggests that an evolving version of the pandemic is an ongoing reality. As it has been said many times, in many ways, this is the new normal. Over these last two years, the adoption of computer vision has accelerated, in part, because of its ability to take on pandemic-related challenges such as monitoring social distancing in public places and ensuring COVID-related cleaning and PPE compliance. Enterprises are also deploying vision AI models to automate and improve analysis of video-driven business processes like defect detection in manufacturing, traceability in supply chains, livestock management in agriculture, task optimization in warehouses, automation in restaurant drive-thrus, and shopper personalization in retail.
Humans will continue to be the intelligence behind a lot of this work, but we will enhance human capabilities with the accuracy of machines with machine learning for work that is prone to human error. Businesses will choose to run operations leveraging cameras to optimize employee placement and assignments to mitigate exposure to physical or mental risks. For example, vision AI confirms and helps enforce masking compliance at the entrances of buildings. It can also detect and alert to workplace hazards to protect workers and reduce workplace injury.
Looking ahead, here are three predictions for 2022:
If 2021 was the year of investments in vision AI, we expect 2022 to be the next phase of industry maturity and consolidation across products. Several specialized product companies will come together to form end-to-end platforms that simplify the application and rapid iteration of vision AI solutions. While the use of video cameras to capture real-world information has been around for years, we’ve optimized the value of this unblinking, consistent data source by integrating visual data with AI intelligence for accurate, real-time, actionable insights.
Second, conversations around ethical AI and ethical data analytics have attracted tremendous attention from the public. We will see greater conversation between AI companies, academia, and policymakers to come to an agreement on an ethical framework for the use of AI. There are two purposes to seeing this through. First, having a common regulatory benchmark for AI will allow businesses to feel confident that they are using AI ethically, and second, it builds accountability for AI companies to minimize bias.
Third, we will continue to be amazed by the innovative use of vision AI. This year, in agriculture alone, cameras and vision AI models were used for everything from identifying individual cows to counting sheep, to tracking animal health, to harvesting robots, to identifying pests and predicting farm production yields.
We expect to see an increased application of vision AI in the world not currently monitored by pre-installed cameras. For example, use cases include installation in the mountains of California to detect early warnings of fire, to cities leveraging vision AI to better assess traffic and congestion conditions.
Vision AI is one of the most exciting areas of technology due to its ability to help businesses and organizations manage and take advantage of the valuable insights hidden within the explosion of images, video, and other rich media. Companies and governments are betting big on Vision AI because of its potential to see beyond the surface; with one caveat, we should commit to developing this technology responsibly using inclusive teams and ethical guidelines.