Mach9 is on a mission to redefine how unstructured 3D data is transformed into maps automatically, enabling faster and more accurate solutions for infrastructure design and asset management. At Trimble Dimensions 2024, we caught up with co-founder Alexander Baikovitz to ask him about the company’s journey, its position in the geospatial ecosystem, and its vision for the future of infrastructure development.
What does Mach9’s mission of accelerating the development of global infrastructure with its geospatial data solutions mean in practice?
Everything that’s designed and built starts with a map. Improving the accuracy and productivity of map-making is a way to develop and build the next generation of infrastructure such as roads, utilities, telecommunications and many other systems that keep the world running.
We’ve created – and are actually defining – a new category of engineering software that we call ‘automated geospatial production’. With AI-powered feature extraction we’re helping our partners make maps up to 96 times faster than manual methods. We also provide really powerful and intuitive quality assurance and quality control workflows to accelerate project delivery. We’re seeing surveying and engineering firms around the world use our solutions on projects ranging from city blocks to an entire subcontinent, helping them make accurate maps faster than ever before.
How did Mach9 become the company it is today?
Many of our early team members come from the autonomous driving industry and Carnegie Mellon. Some of my advisors at Carnegie Mellon were early roboticists, and some came from a civil engineering and survey background. Our early days in infrastructure, working on programmes for Fluor and the US Department of Energy, led us to the core ideas that became Mach9. When we started the company four years ago, we brought together leading AI researchers and engineers from Carnegie Mellon and the autonomous driving industry to transform how computers make digital maps. We started actually building mobile mapping hardware. We learnt first-hand about the challenges in creating point cloud data, and quickly came to understand the true meaning of accuracy and productivity in the surveying and engineering industry.
There are some great organizations like Trimble, RIEGL, Leica, Teledyne and many others that are building the most productive ways to capture data, but not all the industry’s challenges are purely hardware-related. We found a meaningful gap in how people turn the resulting point clouds into the valuable schematics, blueprints and maps that are needed to design, develop and maintain large infrastructure systems and networks. And so about two years ago, we pivoted the company away from hardware towards building Mach9 Digital Surveyor – a software-only solution to address the gap in generating high-fidelity, accurate and fast insights at scale to meet the infrastructure industry’s rising needs for automation.
How is your company embedded in the broader ecosystem of the geospatial industry?
Mach9 Digital Surveyor is interoperable with existing CAD and GIS tools, as well as many other asset management and project delivery tools that our customers and partners rely on to execute on their projects. And the importance of interoperability has been underlined once again here at Trimble Dimensions: it’s essential to connect the powerful data from the hardware that Trimble and other equipment manufacturers to computer-aided design (CAD) software like Bentley MicroStation or Autodesk Civil 3D for design engineering, and geographic information systems (GIS) software like Esri ArcGIS for asset management.
Data interoperability of automated geospatial production software like Mach9 Digital Surveyor with common engineering software systems is critical. By connecting Mach9 maps into downstream infrastructure design and asset management project workflows, we are helping surveyors, engineers and owner/operators to accelerate the development of the next generation of infrastructure around the world.
How does your solution unlock value across various sectors and applications?
Mach9 is able to extract a superhuman amount of detailed information about the physical world. Often, we fit into two complementary systems and workflows: one in the infrastructure design space, and the other in infrastructure asset management. On the design side, we’re helping to automate workflows enabling our partners to create the schematics and as-builts to design major infrastructure systems. On the asset management side, we provide the fastest solutions for processing massive-scale content or data, even covering hundreds of miles, with maps delivered overnight.
Processing that used to take two to four days per mile in some cases can now be done in around ten minutes instead. This is truly transformative and illustrates the huge productivity and efficiency gains that we’re bringing to the industry, enabling the delivery of higher-fidelity information up to 96 times faster, without sacrificing accuracy. Our solution is being used to automate survey applications and production workflows for designing transportation infrastructure, and maintaining assets like signage across entire states, countries and even continents.
What can we expect from Mach9 in the near future?
We are on track to extract hundreds of thousands of miles’ worth of data in 2025. And we see the infrastructure industry’s growing need to continue to optimize how people are collecting this information and how they can use it. Our customers and partners across many critical industries – owners, operators, and engineers in transportation, utilities, rail and other critical infrastructure – are responsible for tens of millions of miles that we’re positioned to help map over the next few years. We are confident that Mach9 will be able to address the industry’s growing need for high-resolution, high-fidelity information at scale. That is what Mach9 Digital Surveyor can help our partners accomplish in effectively every state across the USA, Canada and far beyond.
Alexander Baikovitz co-founded Mach9 in 2021 with the goal of building a platform for automated geospatial production, leveraging advanced AI feature extraction to swiftly transform unstructured 3D data into maps. He previously worked as a Space Technology Research Fellow at NASA Ames Research Center, and gained experience as a robotics researcher at Carnegie Mellon University, USA. He earned both his BSc in Mechanical Engineering and MSc in Robotics from Carnegie Mellon.
