Predictive analytics is one of the emerging areas within innovative industries, and Konica Minolta is investing resources in the research on this topic. In our laboratory, I have been working on a Predictive Analytics for Maintenance project, whose aim is to predict the requests for maintenance of Multifunctional Printers (MFPs) within the organization.
Predictive Maintenance of MFPs is a challenging task from both organizational and technical perspectives. Nevertheless, a system capable to predict the status of the many machines spread all over the world would constitute great value for our company. The first step to set up such a system is collecting signals and information from sensors within MFP devices: at this stage, hundreds of signals are generated. Then raw signals are transformed into a unified format for all models of MFPs and data are stored for:
- preparing machine learning models that predict required maintenance
- predicting the maintenance needed in the following days.
The results of the predictions enable the organization to trigger customer support requests. According to our research, the quality of predictions is determined by the quality of input data and the complexity of the model, whose preparation is among my main research tasks. In our laboratory, we have learnt how to develop a Proof of Concept for such a system in a relatively short time.
However, the details about our solution, will be disclosed within my talk “How to quickly prototype machine learning systems” in the program for the Machine Learning Prague Conference on the 21st -23rd April. MLPrague is the largest conference in Czech Republic for applications of Machine Learning and the organizers have prepared an interesting program with talk and workshops from invited speakers belonging to several of the industrial players in this field such as Facebook, Google, Microsoft and many others.
I am looking forward to take part in the Machine Learning Prague conference and it is an honour to be a speaker there. I will be happy to share with the audience the lessons we have learnt in Konica Minolta Laboratory Europe about solving complex problems for practical applications of the most recent Machine Learning tools. Join us in the Machine Learning Conference on April 22 at 12 am in Lucerna Cinema in Prague.