
Artificial intelligence and machine learning algorithms to detect defects in additive manufacturing by fusing multiple sensor dataAward last edited on: 3/4/2023
Sponsored Program
SBIRAwarding Agency
DOD : NavyTotal Award Amount
$139,926Award Phase
1Solicitation Topic Code
N222-117Principal Investigator
Charles BabbittCompany Information
Phase I
Contract Number: N68335-23-C-0061Start Date: 11/7/2022 Completed: 5/9/2023
Phase I year
2023Phase I Amount
$139,926Benefit:
As the additive manufacturing (AM) technology field has grown, so has the number of different machine producers (OEMs) as well as the range of materials developed for the additive process. The stochastic nature of defect evolution in AM has inhibited its adoption in industries like defense, aerospace, and medicine. While OEMs have expanded the range of materials that can work with their machine and incorporated several sensors to control processing conditions, the data generated by these sensors are underutilized. The project team will seek to develop a machine learning and artificial intelligence (ML/AI) solution that can combine data from multiple sensors for enhanced real-time detection of issues in the LPBF process. Some potential benefits from this project are: Increase machine and user productivity; Improved accuracy in real-time issue detection; Localization of defects for reduced post-processing costs; Improved part quality; Addigurus machine agnostic solution will accelerate the adoption of AM. This solution will help qualify Additive Manufactured parts quicker and save time and money (as much as $150,000 per machine per year) for the Navy, DoD, and commercial users. For instance, numerous operational units and support facilities under the Navy are either not using or have not received appropriate training to use the advanced AM equipment. To leverage the AM capabilities, it is crucial that a solution is developed that utilizes sensory data to provide real-time issue detection. As the Navy pursues enhanced operational readiness and reduced lead times for highly specialized equipment, real-time monitoring will help accelerate the adoption of AM technologies by providing information on the build process and part quality. Some examples where this in situ monitoring technology will be helpful are - - Submarine parts; - Weapons and spare parts; - Repair of ship parts; - Aircraft metal parts; - Aircraft interior parts; - Automotive - low-weight parts; - Consumer products
Keywords:
laser powder bed fusion, laser powder bed fusion, in situ monitoring, additive manufacturing, Real-time monitoring, Machine Learning, Defect Detection, Artificial Intelligence
Phase II
Contract Number: ----------Start Date: 00/00/00 Completed: 00/00/00