#3 Predictive maintenance for a crankshaft manufacturing system

High throughput crankshaft manufacturing system, the principal product of ETXE, is a semi-autonomous
CNC machine that produces an average of 1.000 crankshafts per day. The machine is already IoT-ized and
produces a huge amount of data describing many variables from PLC, CNC and external sensors with
different sampling rates: from 0,5 to 20kHz. In IoTwins these data will be processed and exploited on-line
on the edge to develop real-time predictive maintenance solutions, and will also be sent to the cloud to
accumulate training data (towards optimal maintenance strategies and off-line machine degradation
monitoring) whenever real-time conditions are not required. The fusion of all the data/information could
generate valuable knowledge of the process for the manufacturing line operator to achieve “near future
required” availability of 98%, by integrating with predictive maintenance as a key enabler. Also, note that
ETXE will offer three types real deployment environments for this test-bed, depending on the “maturity
level” of the development: lab, factory-lab, and production line. This testbed is certified by Industrial
Internet Consortium providing part of the infrastructure needed at the three different levels.
The main activities are aimed to develop a new big-data-enabled manufacturing framework that integrates
over the same hardware:
1. sensor fusion at both IoT twins and edge twins;
2. application of predictive maintenance plans for specific parts;
3. the possibility of fully benefiting from advanced big-data visualization techniques for

  1. Provide a three-steps real use-case scenario with industrial pilots, depending on the level of development: lab, factory-lab and production line

  2. Develop predictive maintenance applications to the most critical parts of ETXE machines, concretely X-FLEX machines, such as spindle heads and ball-screws

Involved Partners

#2 Machine tool spindle predictive behaviour
#4 Predictive maintenance and production optimization for closure manufacturing