NOAA Building ‘Digital Twin’ of Earth to Study Climate Change
NOV 14, 2022
Carderock collaborates with warfare centers on Digital Twin prototype
SEP 01, 2021
Naval Surface Warfare Center (NSWC), Carderock Division is part of a new collaboration that involves NSWC’s Port Hueneme Division and Philadelphia Division, along with academia and several small businesses. The collaboration is a Naval Innovation, Science and Engineering (NISE) project.
The goal of the NISE project is to design, develop and demonstrate an autonomous, continuous monitoring system to analyze the health of Hull, Mechanical and Electrical (HM&E) systems aboard the Navy’s Self Defense Test Ship (SDTS), which is stationed in Port Hueneme, California. The SDTS, formerly the USS Paul F. Foster (DD 964), has been repurposed to provide capability as a laboratory in a realistic at-sea environment with the primary mission as being the world’s largest test ship to operate in an unmanned and remote control mode for safe live-fire testing of ship defense systems.
A primary component of the project is the utilization of the SDTS to develop and demonstrate a Navy Digital Twin (NDT) Condition Based Maintenance (CBM) capability for the fleet.
“Through the application of these technologies, this project provides an opportunity for this Navy team to find ways to increase ship maintenance effectiveness by decreasing its downtime and get vessels ready for future missions more efficiently than before,” Carlos Boisselier said, a systems engineer in the SDTS Division at NSWC Port Hueneme Division.
The project duration is planned for three years, with the overarching goal to provide actionable information to decision makers via a prognostic health management system. Authorization for the system installation aboard the SDTS was granted in July 2021.
“We’ve been given three years to work on this project, but we hope to have it completed and to the fleet before then,” Dr. Ben Grisso said, a mechanical engineer in Carderock’s In-Service Ship Structures Branch, who has been working on developing methods to enhance structural hull health for over a decade.
The current system in place to analyze the health of certain submarine rotating and reciprocating machinery ships is to have Sailors walk around their ships with handheld analyzers that collect vibration data, which may then be sent for review by a subject matter expert. This new system will not only allow Sailors to return to their primary duties, but will also provide more frequent and accurate measurements.
“We don’t want to saddle our Sailors with duties that technology can accomplish,” Dr. Michael Robert said, the Technical Project Manager for NDT in Carderock’s Emergent Technology and Signature Analysis Branch. “The current methodologies also result in false positive measurements and information feedback time delays. What we are hoping to achieve will allow the process to be better and faster, with warfighters on the loop as opposed to in the loop.”
The impetus of the project is to ultimately outfit machinery and hull structures on submarines and carriers with sensors that can collect data and utilize machine learning algorithms to determine if the ship is running properly, or to detect and report – and eventually predict – any deviations from expected operational states.
“This equipment utilizes edge computing technology with the idea of computing extracted features, which allows us to collect data,” Sherwood Polter said, an electronics engineer and technical point of contact (TPOC) at NSWC Philadelphia Division. “This information we collect will lead us to a likelihood of a failure mode. The more data we can collect, the better our models will be in the long run.”
NSWC Philadelphia Division’s role in this process is primarily to perform testing and collect data measurements to develop the proper algorithms for the system, as well as extract data for different failure modes.
Digital twins are a virtual representation of complex objects or systems, and as such can be used to predict future operational abilities of ships.
“Our systems will collect a plethora of data, which you can then use to train algorithms that will know whether it is good data or bad data, or trending in a certain direction” Robert said. “The smarter these algorithms become, the systems will begin to be able to make predictions of future HM&E performance.”
Robert, who is a principal investigator for this project, is also Carderock’s TPOC. He has been working with his team to write machine learning algorithms and in turn apply them to NDT CBM, aimed at performing maintenance only when the need arises.
“Essentially, with a ship, if you fix something before it needs to be fixed, then you have increased maintenance costs and decreased operational availability,” Robert said. “If you wait too long to fix things, then you damage the ship and increase maintenance costs, subsequently decreasing operational availability. What we are proposing is an idea to be able to accurately determine exactly when a ship needs to be maintained. NDT CBM tells you before you pull into port what is wrong and what needs to be fixed, which in turn increases operational availability of ships and decreases maintenance costs.”
The collaboration team is taking their development to the SDTS in January 2022 to run tests and collect data to ensure everything works properly.
“Our overarching objective is to mature the technology levels of NDT CBM initiatives to have a prototype that is ready to be installed on unmanned undersea vehicles and submarines,” Robert said.
In order to achieve this objective, development and testing in an operational environment needs to be pursued, which is why they are in need of securing the SDTS.
“We have our initial algorithms written, but they need to be trained on SDTS operational data,” Robert said. “The SDTS is a test bed for the Navy to try out new technologies. While we are aboard the SDTS, we will collect operational data and monitor trends of those data. We can’t just put data measurement and collection systems – enable by machine learning algorithms on the SDTS and walk away — we need to characterize the operational environment for a week or more and then evolve the system over the course of a year or two.”
The project is not exclusive to the Navy — several small businesses and academia, such as Metis Design Corporation, Intelligent Fiber Optic Systems Corporation, Luna Technologies, KCF Technologies and Massachusetts Institute of Technology are also involved.
“Even before Covid, but especially since Covid, we wanted to ensure we included our small business partners,” Grisso said. “By having them help with the development process, it not only helps them stay in business, but also allots us more time to focus on the testing aspect of the technology.”
The end goal is to prove the system successful and to have it to the fleet within the three-year timeframe.
“If we are successful, at the end of our three years we will hope to prove the worth of the digital twin NDT CBM initiatives to the rest of the fleet so that we can have it demonstrated on submarine platforms,” Robert said. “By the end of Fiscal Year 2023, the technology should be mature enough to where it can be transitioned to the fleet. To secure funding for that, we first need to prove its value.”
Elon Musk and top AI researchers call for pause on ‘giant AI experiments’
MAR 29, 2023
Some Writers are Using Artificial Intelligence to Generate Their Articles
OCT 16, 2022
OCT 05, 2022
SEP 21, 2022
SEP 21, 2022