Methods and systems for printed circuit board design based on automatic corrections
In one embodiment, a computing system may access design data of a printed circuit board to be produced by a first manufacturing process. The system may analyze the design data of the printed circuit board using a machine-learning model, wherein the machine-learning model is trained based on X-ray inspection data associated with the first manufacturing process. The system may automatically determine one or more corrections for the design data of the printed circuit board based on the analysis result by the machine-learning model.
1. A method comprising, by a computing system:
accessing design data of a printed circuit board to be produced by a first manufacturing process;
analyzing the design data of the printed circuit board using a machine-learning model, wherein the machine-learning model is trained based on X-ray inspection data associated with the first manufacturing process; and
automatically determining one or more corrections for the design data of the printed circuit board based on the analysis result by the machine-learning model, wherein the one or more corrections adjust one or more parameters associated with the printed circuit board, and wherein the one or more parameters comprise one or more of:
a width of a metal arm or conductive trace,
a length of a metal arm or conductive trace,
a local thickness of conductive trace,
a distance between conductive traces,
a roughness of a conductive trace surface,
a sheet layer thickness,
a distance or spacing between layers,
a corner radius of a conductive trace path,
a location of a plated-through via, or
a dimension associated with a plated-through via.
2. The method of claim 1 , wherein the machine-learning model is trained using a plurality of training samples generated based on the X-ray inspection data, and wherein the X-ray inspection data is associated with printed circuit boards that are manufactured by the first manufacturing process.
3. The method of claim 1 , wherein the machine-learning model is trained using a plurality of pre-labeled training samples.
4. The method of claim 1 , further comprising:
providing the one or more corrections for the design data of the printed circuit board to a design tool used for creating the design data of the printed circuit board.
5. The method of claim 1 , further comprising:
automatically modifying the design data of the printed circuit board based on the one or more corrections; and
providing the modified design data of the printed circuit board to the first manufacturing process, wherein the printed circuit board is manufactured based on the modified design data of the printed circuit board using the first manufacturing process.
6. The method of claim 5 , wherein the one or more corrections adjust one or more parameters associated with an impedance of the printed circuit board, and wherein the impendence of the printed circuit board with the one or more adjusted parameters is independent from layer thickness variations of the printed circuit board.
7. The method of claim 5 , further comprising:
determining a root cause associated with a difference between the design data of the printed circuit board and the printed circuit board manufactured by the first manufacturing process;
generating feedforward information based on the root cause; and
providing the feedforward information to a manufacture tool used in the first manufacturing process to cause the manufacturing tool to adjust one or more parameters associated with the root cause.
8. The method of claim 1 , further comprising:
analyzing the design data of the printed circuit board using a rule-based algorithm comprising a plurality of rules, wherein the one or more corrections are further determined based on one or more of the rules of the rule-based algorithm.
9. The method of claim 8 , wherein one or more rules of the plurality of rules are generated based on pre-determined parameters related to the first manufacturing process.
10. The method of claim 8 , wherein one or more rules of the plurality of rules are generated based on the X-ray inspection data associated with previously manufactured printed circuit boards, and wherein the previously manufactured printed circuit boards are manufactured by the first manufacturing process.
11. The method of claim 8 , further comprising:
determining a difference between previous design data and a corresponding previous printed circuit board manufactured by the first manufacturing process based on the previous design data; and
determining a root cause associated with the difference between the previous design data and the corresponding previous printed circuit board manufactured by the first manufacturing process, wherein one or more rules of the rule-based algorithm are determined based on the root cause.
12. One or more computer-readable non-transitory storage media embodying software that is operable when executed to:
access design data of a printed circuit board to be produced by a first manufacturing process;
analyze the design data of the printed circuit board using a machine-learning model, wherein the machine-learning model is trained based on X-ray inspection data associated with the first manufacturing process; and
automatically determine one or more corrections for the design data of the printed circuit board based on the analysis result by the machine-learning model, wherein the one or more corrections adjust one or more parameters associated with the printed circuit board, and wherein the one or more parameters comprise one or more of:
a width of a metal arm or conductive trace,
a length of a metal arm or conductive trace,
a local thickness of conductive trace,
a distance between conductive traces,
a roughness of a conductive trace surface,
a sheet layer thickness,
a distance or spacing between layers,
a corner radius of a conductive trace path,
a location of a plated-through via, or
a dimension associated with a plated-through via.
13. The media of claim 12 , wherein the machine-learning model is trained using a plurality of training samples generated based on the X-ray inspection data, and wherein the X-ray inspection data is associated with printed circuit boards that are manufactured by the first manufacturing process.
14. The media of claim 12 , wherein the machine-learning model is trained using a plurality of training samples, and wherein the plurality of training samples are generated by manufacturing engineers based on experiential knowledge during a human correction process.
15. The media of claim 12 , further embodying software that is operable when executed to:
providing the one or more corrections for the design data of the printed circuit board to a design tool used for creating the design data of the printed circuit board or to a user of a computing system.
16. A system comprising one or more processors; and one or more computer-readable non-transitory storage media coupled to one or more of the processors and comprising instructions operable when executed by one or more of the processors to cause the system to:
access design data of a printed circuit board to be produced by a first manufacturing process;
analyze the design data of the printed circuit board using a machine-learning model, wherein the machine-learning model is trained based on X-ray inspection data associated with the first manufacturing process; and
automatically determine one or more corrections for the design data of the printed circuit board based on the analysis result by the machine-learning model, wherein the one or more corrections adjust one or more parameters associated with the printed circuit board, and wherein the one or more parameters comprise one or more of:
a width of a metal arm or conductive trace,
a length of a metal arm or conductive trace,
a local thickness of conductive trace,
a distance between conductive traces,
a roughness of a conductive trace surface,
a sheet layer thickness,
a distance or spacing between layers,
a corner radius of a conductive trace path,
a location of a plated-through via, or
a dimension associated with a plated-through via.
17. The system of claim 16 , wherein the machine-learning model is trained using a plurality of training samples generated based on the X-ray inspection data, and wherein the X-ray inspection data is associated with printed circuit boards that are manufactured by the first manufacturing process.
18. The system of claim 16 , wherein the machine-learning model is trained using a plurality of training samples, and wherein the plurality of training samples are generated by manufacturing engineers based on experiential knowledge during a human correction process.
19. The system of claim 16 , wherein the system is further configured to:
providing the one or more corrections for the design data of the printed circuit board to a design tool used for creating the design data of the printed circuit board or to a user of a computing system.
20. The system of claim 16 , wherein the system is further configured to:
automatically modify the design data of the printed circuit board based on the one or more corrections; and
provide the modified design data of the printed circuit board to the first manufacturing process, wherein the printed circuit board is manufactured based on the modified design data of the printed circuit board using the manufacturing process.