With a focus on developing mass-producible, low-cost defense systems, Mach Industries is on a mission to ensure American dominance in the global theater. But to achieve this ambitious goal, they needed a data solution that could keep pace with their rapid iteration cycles and complex engineering challenges.
The Quest for Observability
When JB Rajsky, Mach's Head of GNC & Flight Sciences, joined the company, he saw an opportunity to enhance their data infrastructure. "We were working with CSV files or binary files, loading them into off-the-shelf software for analysis,” Rajsky explained, “While it got the job done, we knew there was room for improvement in terms of efficiency and ease of use.”
This fragmented approach to data management was more than just an inconvenience – it was a significant barrier to progress. "A few weeks ago, we took multiple vehicles out to the desert for flight testing," Rajsky recounted. "We had some interesting results on the first test, and it was nearly impossible to figure out what happened in the field. Not many folks knew how to access the data and provide rapid data review turnaround."
The Build vs. Buy Dilemma
When Mach Industries began to outgrow their existing data infrastructure, they were faced with a critical decision: build an in-house observability solution or partner with an external provider. The choice was not an easy one, as Rajsky, a self-proclaimed fan of vertical integration, initially leaned towards building the solution internally. "If we were to go down the route of using open-source tools to build our own telemetry storage, analytics, report generation type solution, I think that's something that we definitely could do," Rajsky acknowledged. "But whether it makes sense to do right now is a different question."
After careful consideration of the costs, risks, and potential impact on their core competencies, the scales tipped in favor of finding the right partner. "I feel like getting to the capability that Sift has out of the box, even with really good folks, is probably 6 to 12 months, and a team size of 5-10 folks to build, maintain, and scale," he noted. This would be a significant investment for a company of Mach's size, and more importantly, it would divert attention away from their primary focus: manufacturing innovative defense vehicles and developing a robust flight control and simulation stack.
Rajsky's previous experiences had taught him the importance of having a reliable observability solution in place. "I've seen this before and understand the value of having predictions before a flight test. You want to look at the data generated from that flight test, see what lines up, and identify what doesn't. The discrepancies are what you need to fix. Without that, it's really hard to iterate rapidly." Every test flight conducted without easy access to comprehensive data analysis added unnecessary time to Mach's product development cycle.
Risks of Building
Despite the allure of vertical integration, Rajsky recognized the myriad risks associated with building an in-house observability solution. "I'm actually a huge fan of vertical integration, so personally, I would prefer to build this in-house. But I think the major risks are the time to usefulness, as I mentioned earlier. It's very beneficial for us to have this tool now rather than in 6 months as we continue testing at the component level and with full up flight testing in the field. There's a risk of progress delays and the associated cost in time."
Maintenance costs were another looming concern. Rushing the development process or allocating insufficient resources could result in a system that buckles under the strain of Mach's growing data needs. "The other risk is maintenance cost, which could be quite high—a substantial recurring cost in terms of both headcount and resources. If we don't implement the right scalable solution now because we rushed and didn't allocate enough resources, we might face significant issues when we try to scale. This could set us back significantly and result in technical debt from building something that doesn't scale effectively to our needs as the company grows."
Moreover, Rajsky understood that observability was not currently Mach's core competency. Diverting their attention to building and maintaining an observability solution would be an unnecessary distraction from their main objectives. Partnering with a dedicated provider like Sift, whose sole focus is on delivering best-in-class observability tools, would allow Mach to benefit from their expertise and continuous improvements driven by a diverse customer base.
Read more about the risks of building in house in Sift's Aerospace Observability Report.
Unlocking Insights with Sift
With Sift's observability platform, Mach has been able to streamline its data management and analysis processes significantly. "The UI is very user-friendly and easy to use," Rajsky noted. "It has certainly streamlined the data review process for the flights that we've done with it so far."
Sift's ability to ingest and normalize data from a wide range of sources has been particularly valuable for Mach. "I know Sift’s back-end is Apache Parquet," Rajsky mentioned, "and I think that will also potentially pay dividends for us in the future if all the data is in one source. It's a schema that we understand, and we can do bulk processing on a bunch of historical factory testing or flight testing data."
Sift has certainly streamlined the data review process for the flights that we've done with it so far.
JB Rajsky
Head of GNC & Flight Sciences, Mach
But Sift's impact goes beyond just data aggregation and visualization. With features like real-time anomaly detection, and predictive maintenance, Sift is helping Mach proactively identify and resolve issues before they lead to costly failures.
A Foundation for Scale
Partnering with Sift will yield significant, tangible benefits for Mach Industries, particularly in their ability to scale manufacturing and mass-produce their innovative defense products. "The impact on Mach Industries is significant," Rajsky emphasized. "Securing contracts from the DOD and potentially foreign customers hinges on our ability to conduct thorough design work, flight testing, and analysis efficiently. Streamlining our design cycle with robust data management is crucial."
Securing contracts from the DOD and potentially foreign customers hinges on our ability to conduct thorough design work, flight testing, and analysis efficiently. Streamlining our design cycle with robust data management is crucial.
JB Rajsky
Head of GNC & Flight Sciences, Mach
Sift's ability to handle the massive volumes of data generated by Mach's vehicles, both in testing and in production, is essential to achieving this goal. With Sift's platform, Mach can continue to iterate and innovate at a rapid pace, without worrying about outgrowing their data infrastructure.
Tangible ROI
With Sift's unified observability platform at the heart of their operations, Mach has been able to begin optimizing every stage of their "design-build-test-fly-iterate" cycle. "Enhancing the speed and efficiency of this iteration loop, particularly through streamlined data ingestion and reporting, is immensely valuable to us," Rajsky noted. This optimization has led to accelerated engineering cycles, improved product reliability, and a solid foundation for large-scale production.
Sift's platform has given Mach a competitive edge, enabling them to present a compelling, data-driven case to potential customers and regulators alike.
The return on investment extends beyond the engineering process itself. "Whether bidding for government contracts or undergoing certification processes, having easily accessible flight test data for review, approval, and report generation is pivotal," Rajsky pointed out. Sift's platform has given Mach a competitive edge, enabling them to present a compelling, data-driven case to potential customers and regulators alike.
A Partnership Built on Trust and Innovation
Beyond the measurable ROI, what sets Mach's partnership with Sift apart is the trust, transparency, and shared commitment to innovation that underpin their relationship. "Sift was really open with how their backend works and their technology roadmap," Rajsky noted, "and that was really refreshing for us to see that we could be relatively integrated with their product development as we make progress."
This collaborative approach has been instrumental in helping Mach foster a culture of data-driven decision making. "Personally, I'm focused on fostering a culture that values analyzing data and results from test flights," Rajsky shared. "Based on my previous experience, having clear expectations and thorough post-flight analysis accelerates the engineering cycle significantly. Implementing this now will pay dividends as we expand our product line and gain traction with current offerings."
The team at Sift has tackled this problem directly before, and they are able to leverage a lot of that expertise and past knowledge and past experience to help us.
JB Rajsky
Head of GNC & Flight Sciences, Mach
In Sift, Mach found a partner uniquely qualified to meet their observability needs. With a team deeply experienced in the challenges of complex hardware systems and a platform purpose-built for the task, Sift stood out from the competition. "It's the team," Rajsky said. "The team at Sift has tackled this problem directly before, and they are able to leverage a lot of that expertise and past knowledge and past experience to help us."
As Mach continues to push the boundaries of defense manufacturing, they do so with the confidence that Sift will be right there with them, evolving and adapting to meet their ever-changing needs. This partnership is not just about technology — it's about a shared vision for the future of the industry and a commitment to driving innovation together.
