From Scientists to Satellites: LMSS Tackles Disparate Life Cycle Data with a Neo4j Product 360 System
The ChallengeLMSS was in dire need of a solution to support digitizing and integrating all of their processes and data across the entire lifecycle of products. With a wide expanse of disparate data they had no flow across multiple systems.
Ann Grubbs, Chief Data Engineer for LMSS, said they’d built a few interfaces to connect data, but that “it cost us a kazillion dollars to build the interface between our data storage systems, and it’s not very scalable if you want to look at the entire life cycle of a product.”
Most equipment LMSS builds has a very long development life cycle. From engineering to launch, every facet of the manufacturing life cycle correlates with and affects each other. Redesigns done today could have a big impact on something that’s being put together years later.
As technology moved forward, LMSS amassed far more data than a human could ever understand or manage.
“All I can tell you is there are hundreds, maybe thousands of data systems, and tens of thousands of datasets,” Grubbs said. “We create a lot of data around here.” Though a lot of information was residing in the heads of very smart people, tribal knowledge was too unreliable. Instead of a customer 360 solution – because they only have one customer, the government – they needed a product 360 solution.
The Solution“We looked at the problem and had a diagram that had circles coming off from lines, and it was a representation of our graph omen,” Grubbs said.
Her team went searching for graph databases and found Neo4j. With Neo4j, LMSS stored their datasets in a graph of connected data. With reporting and analytics capabilities, they’re able to easily see how their data fits together across a product life cycle.
Before using graph technology, it could take weeks to query all of the disparate systems to find an answer to an analyst or manager’s question. The common methodology was to assign someone to manually connect the dots. By using a graph database to query data connections, however, the finding an answer to a question became much more efficient.
“Neo4j’s graph database created a map of our products,” added Grubbs. “’The DNA of our products’ is what we like to say.”
The LMSS team has a polyglot mantra, and that’s been a huge part of Neo4j’s successful implementation. Neo4j guides the application to the appropriate legacy system to drill down, bit by bit, and connect all data silos – like a nervous system. Because the massive architecture of their information “map” is built around Neo4j, they’re set to scale.
LMSS understood early on they could use the graph for numerous things in a practical and valuable way. All the way up to the CEO, different departments want data connections that make sense for their objectives.
“We’re rewriting our whole framework so we can blend this broad data with our deep data,” Grubbs said. “We have all kinds of business cases lined up and ready to go.”