In-depth analysis tagged Market Analysis — covering PLM history, vendor strategy, and the technical decisions reshaping engineering software.
4 articles
ThreadMoat counted 2,223 verified customers of industrial software startups. Read from the practitioner's chair, the numbers say the evaluation question is never 'do we have a gap' but 'does the module we already own actually have the capability'. Here is how to use the data when a startup lands on your desk.
ThreadMoat scored 75 engineering-software startups that reached an exit and separated the acquisitions from the shutdowns. Technology was the weakest predictor; growth and funding efficiency were the strongest. If you are the buyer, that means your due diligence is weighted on the wrong slide.
A cutting tool maker and a metrology giant own more CAM software than any PLM vendor. The same incumbents are funding, shipping and buying from the 65+ AI-native startups in the ThreadMoat database. Everyone is converging on one layer. I call it Adaptive Manufacturing, and the M&A map says I am not wrong.
600 startups. 10 unicorns. $15.7 billion in venture capital. A parallel engineering software industry is shipping order-of-magnitude workflow improvements while Autodesk, Siemens, PTC, and Dassault stall. Where is the market heading?
Adaptive Manufacturing is production that senses its own state and changes what it does next. It closes the loop between measurement and action on the shop floor: measure the part and adjust the toolpath, detect tool wear and re-plan the operation, feed inspection results back upstream before the next part is cut. The term is process-agnostic, covering subtractive machining, additive manufacturing and the hybrid of the two, and it describes a behavior rather than an infrastructure. It is distinct from Industry 4.0 (a connectivity ambition), smart manufacturing (a marketing adjective) and closed-loop machining (a single feedback circuit). Adaptive Manufacturing is also the name of the ThreadMoat market map category for the 65+ AI-native startups building this sense-decide-act layer on top of existing CAM, CNC and metrology systems.
Software designed from inception around autonomous AI agents that can read, reason about, and act on domain data. Not retrofitted AI (chatbot on the sidebar); deeply integrated into the system's core.
A parallel industry of startups solving problems that incumbents have failed to address. The 'shadow' part means they operate outside the incumbents' sphere of influence—different investors, different customers, different technology stack—but address the same end-user problems.
STEP-NC, formally ISO 10303 AP238, is a standard that extends the STEP product data model to machining. Instead of the 1950s-era G-code emitted by a post-processor, STEP-NC carries associative, machine-portable manufacturing programs that reference the part geometry, features, tolerances and operations directly, so the same program can in principle run on any compliant controller without a machine-specific post-processor.
A toolpath kernel is a licensable software component that computes cutter paths (roughing, finishing, multi-axis, simulation) which CAM vendors embed inside their own products, in the same way CAD vendors license a geometry kernel such as Parasolid or ACIS. ModuleWorks of Aachen is the dominant toolpath kernel supplier; its investors include DMG Mori, Mitsubishi Electric, Kennametal and Autodesk. MachineWorks (Sheffield) is the other significant independent, focused on simulation and verification.