Apparatus and method for in-manufacturing evaluation of structural and material properties of fasteners using machine learning
An apparatus and method for detecting structural and material defects in a fastener driven during a manufacturing process includes a driving tool capable of recording an angle-torque trace during the driving of the fastener and a machine learning engine operably connected to the driving tool for analyzing the recorded angle-torque trace. The machine learning engine can be provided with a number of sample angle-torque traces from sample fasteners and can self-determine a stored trace including tolerances for acceptable angle-torque trace data from the samples in an unsupervised learning process or protocol without the need for defined anomalous and non-anomalous samples being provided to the machine learning engine. Using the self-defined stored trace and acceptable tolerances, the machine learning engine can analyze attributes of subsequently recorded angle-torque traces to ascertain whether the attributes of the recorded angle-torque traces indicate anomalies within the fastener identified by the recorded trace.
1. A method for automatically detecting anomalies in the structure or material forming a fastener utilized in the assembly of a medical device, the method comprising the steps of:
a. providing an apparatus comprising:
i. a driving tool capable of recording an angle torque trace for a fastener driven by the tool to form a recorded trace; and
ii. a processor operably connected to the driving tool and including an machine learning engine stored on the processor;
b. creating a stored angle torque trace for comparison with the recorded trace components with the machine learning engine;
c. driving a fastener to secure components of a medical device with the driving tool;
d. recording data forming the angle-torque trace while driving the fastener to form the recorded trace; and
e. comparing the stored angle-torque trace c with the recorded trace to detect anomalies,
wherein the step of creating the stored trace comprises:
i. providing a number of sample angle-torque traces to the machine learning engine;
ii. conducting an unsupervised learning protocol with the machine learning engine on the sample angle-torque traces; and
iii. defining attributes for the stored trace based on anomalies determined by the machine learning engine within the sample traces.
2. The method of claim 1 wherein the step of providing the number of sample angle-torque traces to the machine learning engine does not include providing examples of pre-determined anomalous and non-anomalous angle torque traces to the machine learning engine.
3. The method of claim 1 wherein the step of defining attributes for the stored trace comprises creating a tolerance ranges for one or more attributes of the stored traces.
4. The method of claim 3 wherein the tolerance range defines a range for a torque value attribute of the stored trace.
5. The method of claim 4 wherein the torque value attribute is a final torque value.
6. The method of claim 4 wherein the torque value attribute is a run down torque value.
7. The method of claim 4 wherein the torque value attribute is a run down angle value.
8. The method of claim 4 wherein the torque value attribute is a total run down value.
9. The method of claim 8 wherein the recorded trace is wirelessly transmitted to the processor.
10. The method of claim 3 wherein the tolerance range defines a range for a slope attribute of the recorded trace.
11. The method of claim 10 wherein the slope attribute is a final torque slope.
12. The method of claim 1 wherein the step of comparing the stored trace with the recorded trace comprises:
a. selecting a number of attributes of one of the recorded trace or the stored trace to compare with the attributes of the other of the recorded trace or the stored trace; and
b. comparing the selected attributes to detect anomalies in the recorded trace.
13. The method of claim 12 wherein the step of selecting the number of attributes is performed by the machine learning engine.
14. The method of claim 1 further comprising the step of transmitting the recorded trace to the processor after driving the fastener.
15. The method of claim 1 wherein the steps of recording the data to form the recorded trace and comparing the recorded trace to the stored trace are simultaneous.
16. The method of claim 1 further comprising the step of grouping the recorded traces by fastener type after recording the angle-torque trace.
17. The method of claim 1 further comprising the step of removing a fastener in which anomalies are detected after comparing the recorded trace to the stored trace.