JAPAN VIDEO REVIEW · MACHINE BUILDER AI
When a Machine Builder Builds Its Own AI: Nakamura-Tome's "3D Smart Pro AI" NC Auto-Generation
01What the two official videos are about
Nakamura-Tome, founded in 1949, is one of Japan's flagship builders of turn-mill and multi-axis turning machines. In May 2023 its official channel posted a video with a strikingly blunt title: "[Job time cut by more than 6 hours!?] A big introduction to the NC auto-generation software '3D Smart Pro AI'!"[1].
The sequel, half a year later, spells out its target audience even more clearly: "Even beginners can easily produce NC programs!"—together the two videos have racked up roughly 17,000 views, a solid result for a machine builder's official channel[2]. Moving the pitch from "fast" to "anyone can do it" maps precisely onto a shop-floor problem shared by Japan and Taiwan alike: too few people can write programs for multi-tasking machines.
02Why turn-mill NC programs are especially hard to write
A turn-mill machine combines the capabilities of a lathe and a mill in a single machine, finishing multiple operations in one setup—and the price for that is a multiplying of programming complexity:
- Switching between turning and milling mindsets: within one program you have to switch between turning coordinate logic and milling coordinate logic, and the combinations of C, Y, and B axes are far more complex than on a three-axis mill.
- Multi-turret synchronization: on machines where upper and lower turrets move at once, you also have to plan the timing and clearance of two sets of tools.
- High collision risk: turrets, tailstock, main and sub spindles, and the workpiece all dance together in a tight space—one wrong step in the sequence and you crash the machine.
Academic reviews of CNC intelligence pointed this out long ago: the more capable the machine, the higher the knowledge barrier for programming and process planning—which is exactly why "more open, interoperable, and intelligent" control and programming technologies keep being researched[3]. Research on automated process planning (CAPP) also lists "multi-tasking integrated machines" as one of the scenarios where automation delivers the greatest benefit—and is the hardest nut to crack[4]. Nakamura-Tome putting AI here is aiming straight at the pain point.
03What it means for a machine builder to build its own AI programming
What's notable isn't just the software itself, but who made it. A machine builder's business model is selling machines; when it invests resources in developing programming AI and shoots a video to push it, that tells you it has reached a conclusion from the customer side: the machine isn't the bottleneck—the program is. You can sell the fastest multi-tasking machine there is, but if the customer can't write the program, that capacity goes to waste—this is the Japanese version of the same "orders stuck at CAM" we covered in the US piece, The CAM engineer talent shortage.
"Job time cut by more than 6 hours" is a claim from the official video title; the number can't be independently verified, but its direction matches AI-programming field tests around the world: the rule-heavy prep work of programming is exactly the segment where AI's compression is most visible (for US third-party testing, see The TITANS of CNC case study). And the "beginners can use it" pitch points straight at skills transfer—letting less-experienced operators produce usable programs while senior staff focus on review and the hard jobs.
04A Taiwan shop's view: proprietary software vs. a neutral prep chain
If you're running Nakamura-Tome machines, staying inside the machine's own ecosystem is usually a reasonable choice—the builder knows its own controller and mechanics best. But the norm for Taiwan's job shops is a mixed multi-brand floor: one brand of lathe, two brands of mill, yet another brand of turn-mill machine—learn a separate proprietary software for each and your knowledge gets sliced into several pieces.
That is where a "machine-neutral" prep chain earns its keep: BestAI CAM starts from the customer's 2D drawing—AI reads the drawing, builds the 3D model, and generates G-code against your in-house tool library, controller, and travel/spindle-speed limits, then runs 3D cutting simulation and independent AI dimensional cross-checking, with people handling tolerances, datums, and the pre-machining sign-off (for the full workflow see The complete guide to CNC auto-programming). Machine-specific AI and a neutral prep chain aren't mutually exclusive: the former digs deep into optimizing a single machine, the latter unifies the front-end "drawing to program" flow—for a multi-brand shop, pairing the two keeps knowledge from being locked into any single ecosystem.
05FAQ
Can I trust the "more than 6 hours saved" number?
This is a claim from the title of Nakamura-Tome's official video—a vendor's own statement that can't be independently verified, so cite it with its source. That said, the general direction—AI heavily compressing the rule-heavy prep work of programming—matches third-party field tests such as those from TITANS of CNC and Modern Machine Shop in the US. The real number for your shop only comes from timing your own parts before and after adoption.
Why would a machine builder develop its own AI programming software?
Because the customer's bottleneck has moved from the machine to the program: the more capable a multi-tasking machine is, the higher the programming barrier, the fewer people who can write for it, and the less of that machine's capacity gets used. A machine builder investing in AI programming is effectively admitting that the program is the bottleneck—which is exactly why the two selling points in the videos are "saves time" and "beginners can use it."
My shop runs machines from several brands—whose AI should I use?
Pair two layers: leave deep single-machine optimization (vibration suppression, on-machine diagnostics) to the machine builder's ecosystem; unify the front-end prep chain—drawing in, model, code out, simulate, verify—with a machine-neutral tool that outputs the right program for each machine's controller and limits. That way the knowledge accumulates in your process instead of being scattered across each vendor's software.
Get notified when new articles and video reviews go live—no inbox flooding, one-click unsubscribe.
MACHINE-NEUTRAL AI PREP
Mixed multi-brand shop? Unify the front end with one neutral AI prep chain
Read the drawing and model from a 2D drawing, generate code against each machine's tool library and controller, simulate and verify dimensions—try it once with your own drawing.
Talk to an onboarding consultant Training courses📋 Download the free "CNC Outsourcing Checklist" (printable) (中文)
06References
- 中村留精密工業公式チャンネル（YouTube）. 【作業時間6時間超減！？】NCプログラムを自動で作成！プログラム作成ソフトウェア「3D Smart Pro AI」を大紹介！ (published 2023-05-16). youtube.com/watch?v=eQGB-HxEVs4
- 中村留精密工業公式チャンネル（YouTube）. 初心者でも簡単にNCプログラム作成！「3D Smart Pro AI」を大紹介！【わたくらコンビ】 (published 2023-12-01). youtube.com/watch?v=Uzu2GUUI_OE
- Xu, X. W., & Newman, S. T. (2006). Making CNC machine tools more open, interoperable and intelligent — a review of the technologies. Computers in Industry, 57(2), 141–152.
- Xu, X., Wang, L., & Newman, S. T. (2011). Computer-aided process planning — A critical review of recent developments and future trends. International Journal of Computer Integrated Manufacturing, 24(1), 1–31.