FUTURE MANUFACTURING
Sources : BOMAG
An AI-based field service app provides customers and service staff with the right solution to existing faults and problems within seconds .
spare parts that will be needed can then be automatically shipped in advance , or the appropriate components can be replaced on site before problems arise . As a result , there is no downtime , and the customers ’ overall equipment efficiency is assured .
Knowledge graphs – thinking like a person
When there are acute problems with machines , parts or components , service employees must quickly find the right solution . But since knowledge is often located in information silos ( such as systems for enterprise resource planning , customer relationship management or product information management , or in e-mails or spare parts catalogs ) and various unstructured formats ( technical documentation , speci- fication sheets , recall notices , service reports ), many systems come up against their limits in these situations . This happens because the systems frequently operate with only precise search terms have no “ understanding ” of the context of the service issue or of the machine or customer . This challenge can be overcome with an AI technique that can intelligently link up all of the existing data and information : knowledge graphs .
Knowledge graphs operate the way people think : they are capable of uncovering the complex and varied relationships and connections among machines , components and parts . Knowledge graphs are not just structures that take in data and return it as output ; they represent the entire logic of machines , product portfolios or industrial installations in a formal digital twin .
Application suggests solutions
Knowledge graphs work their way through different scenarios and are capable of determining what effects a fault can have . They indicate whether the machine has to be stopped and the work interrupted , and what can happen in the worst case . In doing so , the application takes into account all possible parameters , such as the frequency of damage in the past , weights , forces , likely heat generation and much more . In this way , knowledge graphs identify the probability and severity of possible damages . They understand the interrelationships between problems and make relevant suggestions to solve the issues involved . Even if no definite solution is available for the exact problem described , the knowledge graph will still present the service engineer with proposed solutions to similar
28