Digital Intelligent Assistant for Predictive Maintenance as a Response to Demanding Employee Skill Requirements
An EIT-Manufacturing Project
Barriers in predictive maintenance
Implementing predictive maintenance (PdM) in production is a challenging, data-intensive task that requires data science experts, who are capable to generate company specific maintenance strategies. There is intense competition between companies to hire these highly skilled experts. Their cost is an important barrier that limits the adoption of PdM. Manufacturers realize PdM with supportive software.
A digital assistant for predictive maintenance
DIAMOND used artificial intelligence to support less skilled employees during tasks, such as failure diagnosis, forecasting, and maintenance planning. Employees are be able to have a conversation with the assistant about machines and processes. They may use it during meetings, planning, and shop floor tasks where it will provide information about machine health, components, and maintenance events.
Applied Open Source technology
DIAMOND applied and modified the Open Source voice assistant Mycroft. It is privacy-focused, which makes it better suited for industrial environments with high privacy requirements.
Two challenging use cases
We focused on maintenance activities in white goods and medical products manufacturing. Two industry partners provided us their requirements and the setup to demonstrate our solution.
EIT-Manufacturing learning nuggets
DIAMOND received funding from EIT-Manufacturing. We created 5 learning nuggets for their learning platform describing how to design a simple digital assistant that can support predictive maintenance.
BIBA – Bremer Institut für Produktion und Logistik GmbH
Delft University of Technology
Whirlpool EMEA S.p.A.
Stryker Trauma GmbH
This project has received funding from the European Institute of Innovation and Technology (EIT), a body of the European Union, under the Horizon 2020, the EU Framework Programme for Research and Innovation.