Pascal Rossignol/Reuters
- Amazon AWS is launching a series of industrial monitoring tools that use AI to help companies identify potential equipment failures, productivity bottlenecks, and worker safety violations.
- The new tools, previously reported by Business Insider, signal Amazon's growing ambitions in the industrial cloud space.
- One of the tools uses artificial intelligence to analyze video footage and look for "anomalies" in industrial processes, raising concerns given reports of Amazon's history of spying on workers and using racially and gender biased AI.
- An Amazon spokesperson told Business Insider the tool doesn't currently have "pre-packaged facial recognition capabilities," and that AWS "thoroughly tests and audits" its algorithms and uses diverse training data.
- Visit Business Insider's homepage for more stories.
Amazon AWS is launching a series of industrial machine learning tools aimed at helping companies use artificial intelligence to monitor factory workers and machines, the company announced in a press release Tuesday.
The suite of tools, which Business Insider's Eugene Kim previously reported, use predictive analytics to identify potential equipment failures and product defects, worker safety and compliance violations, and areas where they may be able to improve productivity, among other industrial process improvements.
Tuesday's launch reflects Amazon's growing ambition in what is commonly referred to as the "industrial internet of things," which Grand View Research estimates will be worth $949 billion by 2025 and is growing at 30% annually.
"Organizations would like to use the cloud and machine learning to help them automate processes and augment human capabilities across their operations, but building these systems can be error prone, complex, time consuming, and expensive," Swami Sivasubramanian, vice president of machine learning for AWS, said in the release. Amazon's new tools are "easy to install, deploy, and get up and running quickly and that connect the cloud to the edge."
The tools include: Monitron and Lookout for Equipment, sensors and software that predict when machines might fail or need maintenance; and Panorama Appliance, Panorama SDK, and Lookout for Vision, which enable companies to automatically detect a variety of "anomalies" in video footage.
While Panorama and Lookout for Vision are primarily focused on industrial use cases, they also raise concerns about firms' growing reliance on workplace surveillance as well as algorithmic bias, particularly given Amazon's history with regards to both topics.
Vice News reported last week that Amazon hired operatives from a notorious union-busting spy agency to monitor warehouse workers, unions and social movements, and in September, that Amazon also spied on US workers via private social media groups. In May, a longtime AWS vice president quit over the company's treatment of workers during the pandemic.
Amazon also has faced criticism from researchers who said they found evidence of racial and gender bias in its facial recognition product, Rekognition, as well as an AI recruiting tool it was developing. Amazon said in June it would ban police from using Rekognition for one year amid backlash from anti-policy-brutality advocates.
An Amazon spokesperson told Business Insider that "AWS Panorama does not include any pre-packaged facial recognition capabilities."
"AWS thoroughly tests and audits computer vision service results for accuracy whenever there are any changes to a service, and any regression in quality will result in a rejection of the changes to the system," the spokesperson said, adding that Amazon constantly updates its services and uses "diverse training data."
from Business Insider https://ift.tt/2VqAJCk
No comments:
Post a Comment