Easy recording of status conditions on production machines and measurement of order-related energy consumption without interfering with the control electronics.
Almost all manufacturing companies have a heterogeneous machine park with a wide variety of control systems (PLCs). While the machine status can usually be queried at the machine terminals, production management often lacks an overall overview.
Although there are already many technical options for recording machine statuses, many of them are complex, cost-intensive, and different for each machine.
Our solution is a simple, cost-effective, and independent recording of the power flow on a machine. With appropriate hardware, the flowing current is measured every second and transmitted to a server via a Wi-Fi-enabled microprocessor. This processor sends its own ID, allowing signals to be uniquely assigned to a machine.
By recording the second-by-second current flow at individual machines, the obtained values can be evaluated in six categories:
- Machine status (production/downtime/cycle rate)
- Time evaluation of machine usage (OEE key figures)
- Time evaluation of an order
- Machine energy consumption
- Process quality control
- Predictive maintenance
Advantages:
- No intervention in the control system is necessary, and therefore no risk of a warranty breach on the machine.
- Cost-effective and fast method for MDE
- Self-explanatory software application without extensive training measures.
- Minimal internal resources required for implementation
- Individual for each company
- For all machines, regardless of the control system
- Application possible: stand-alone or in conjunction with MES or PPS.
- No regulatory requirements for use in medical technology manufacturing companies.
Hardware:
A Raspberry Pi serves as the basis for the hardware component. A Wi-Fi-enabled microprocessor extends the Raspberry Pi with interfaces for voltage measurement and contactless current measurement. Together with the current transformers, it becomes a fully-fledged smart meter that monitors and records power consumption and generation, and makes it available via the network or Wi-Fi.
A corresponding trial installation has already taken place at a DOQ reference customer and has been providing test data since October 2022.
Software:
DOQ-MES offers an MDE interface that can display the received data as status states. For this purpose, the data is assigned to a resource, i.e., the controller sends an ID that is uniquely assigned to a resource stored in the system.
For customers without DOQ-MES, a separate interface is provided with which the order data can be linked to the machine. This can be done through manual entry or via an interface to the existing ERP system.
The data sent from the controller to a cloud application via a REST service is processed in the background. The software uses an AI algorithm for pattern recognition (current master’s thesis with the Universities of Ulm and Erlangen), which, as a live application, recognizes the pattern curve of a manufactured part from the sent data after a certain period of time. Ideally, an order for 50 parts produces a graph with 50 identical curves. The time axis of an individual curve then corresponds to the piece time. If the curve remains close to zero on the y-axis, the machine is in standby mode and is not producing. The results are transferred to the dashboard interface at specified time intervals.
An increased current measurement during cycles can indicate tool wear. Predictive maintenance/tool changes can therefore be planned if value limits are set for the deviations.
The resulting measurement data can also be used as evidence of consistent quality during product manufacturing.
