How the industrial sector adopts AI technologies

How the industrial sector adopts AI technologies
The industrial segment is embracing new technologies and this is evidenced by numerous launches of AI solutions tailor-made for manufacturers, according to The Industrial Revolution: The Top Trends and Takeaways from Hannover Messe 2019 published by ABI Research.
 
“Companies are not hunkering down but are aggressively pursuing technology as an avenue to increase productivity and drive profit margins,” ABI says.
 
The automotive vertical is leading the push toward digital transformation, says Stuart Carlaw, Chief Research Officer at ABI Research. “This segment is by far the most advanced in terms of automation… as the auto industry starts implementing, the tier 1 and 2 providers are forced into adopting the same digital platforms and then that snowballs throughout the value chain.”
 
Another trend that Carlaw noticed is that companies are now more receptive to cloud technology. “Vendors begin to adopt blended approaches with some on-prem edge capability and some cloud capability that allows for the best of both worlds.” Last year using cloud for mission critical data was “a big no-no,” but this is beginning to chance, Carlaw noted.
 
Artificial intelligence is also a focus among the industrial sector. ABI observed the adoption of AI is growing.
 
While the industrial market was previously dominated by technology firms when it comes to offering AI solutions, today it is traditional industrial players that are leading the launches, said ABI Research.
 
Traditional industrial players include HARTING, Weidemueller, ZF Group, SEW-EURODRIVE and Konica Minolta. “These companies’ core product lines are grounded in the industrial space; they’re focused only on building solutions industrial manufacturing companies are prepared to implement. Consequently, their shift toward AI indicates industrial companies are showing a considerable appetite to mature and embrace AI technology,” ABI says.
 
ABI listed some advanced industrial solutions showcased at Hannover Messe 2019. Weidemueller presented an analytics platform that uses machine learning to support industrial automation.
 
HARTING, a major vendor of industrial cables, has launched an IoT gateway called MICA which provides connectivity and machine learning inference at the edge.
 
ZF Group is working with Nvidia to develop an edge computer platform called ZF ProAI for vehicle automation.
 
SEW-EURODRIVE is one of the first companies in the world to adopt a system of collaborative robotics on its factory floor in Baden-Württemberg, Germany.
 
Konica Minolta has launched an AI platform for image, sound and data analytics for industrial automation, predictive maintenance and quality control.

AI can’t scale in the industrial sector

So far AI still has its limits in industrial applications. One of them is scalability. Scalable AI solutions and applications are tough to come by in the industrial space, according to CraftWorks, a software company that builds custom AI solutions for industrial clients.
 
AI solutions are normally deployed as custom implementations and tailored to a specific application or customer. “It is challenging to train general models that will translate between different clients,” ABI says.
 
Also, clients are often also part of the problem because they don’t want to share their data or contribute fearing their competitors may get the edge. As a result, vendors focus on developing AI platform engines on which their industrial clients can build their own models, as opposed to relying on specific AI solutions.


Share to:
Comments ( 0 )