Analysis Adding robots or automating machines in old factories isn't as easy as it sounds. Retrofitting factories with new technologies for machines and robots to operate in sync requires a new system architecture.
Many tech companies are trying to fill that gap with hardware and software. Cloud companies, chip makers, software companies and telecom providers are spitballing topologies for machines to communicate directly and analyze data.
On Tuesday Intel said it was going the Linux way for system upgrades in factories. Chipzilla has worked with Red Hat on a blueprint that provides a more granular coordination of machines' operation between facilities, which is now being made possible on the faster 5G networks.
The base of the system architecture is built around machine-to-machine communications between robots, actuators, machines and I/O devices on the shop floor. Manufacturers rely heavily on artificial intelligence to automate machines and boost output, in which case data needs to be sent through a series of steps to higher performing servers that then parse the data.
Raw data gathered through a series of edge computing devices with Intel chips optimizes data flow by cutting off irrelevant data, and making sure only relevant information flows upstream to the on-site data for AI models or analytics. The data transport is orchestrated with Red Hat's OpenShift Kubernetes and virtualization platforms operating at various levels.
The data then flows up to data centers on-site or remotely where machine learning can analyze the data. The results can be spit back out directly to machines and robots, which can then adjust operations.
The architecture is designed to operate in individual factories, or across a network of manufacturing sites.
Google recently announced its Google Distributed Cloud initiative, where it will offload some processing of information on the edge via hardware certified by the company. The edge device could sit on edge devices hosted by Google, telecom providers or customers. This could be relevant to manufacturers who would want to use Google's cloud service for AI or other applications.
But Intel's and Red Hat's one-size-fits-all factory system architecture may not work for all manufacturers. Some companies cases are building custom technology blueprints to meet their specific needs.
- Not keen on a 5G mast in your street? At least it'd be harder for crackpots to burn down 'a flying cell tower in orbit'
- Sloppy string sanitization sabotages system security of millions of Java-powered 3G IoT kit: Patch me if you can
- Top IoT M2M module shipper for 2017 was China's Simcom. Who's surprised?
- Good luck saying 'Sorry I'm late, I had to update my car's firmware'
The companies are testing autonomous vehicles communicating directly inside factories with each other on the 5G network, which is built within the confines of the factory and not connected with the public network.
In both cases, they are cutting off AI chips and components that helped the vehicles communicate with each other to avoid collisions. Instead, the automated vehicles have a 5G network chip sending back data to an internal cloud, which is mapping out navigation data and directing robot movement. Siemens is equipping factories with 5G small cells and routers and is using the vehicles to transport materials between machines. ®