That's pretty interesting! Is that programmable via CAM software with post-processor, or is it more involved?

I'm certainly not set on this giant 5-axis with drill bank sort of thing, but I seems like it would be the most capable. That said, it does seem to have a fairly low Z-axis, and some of the videos I've seen seem pretty slow, but it's hard to tell whether things are optimized or not. I did realize at the IWF woodworking show that there were low end 3-axis machines priced at a level where I could buy 10 of them for the price I'm looking at here, and parallel production is where speed comes from. That said, I had a similar realization back when I had our first laser with an 18x24" work area, but the 48x96" laser with 3X the power has been far more capable than 6-8 of those lasers would have been. The thought at the time was surrounding paper cutting for the print industry, and we have since gotten galvo lasers where the beam is steered by mirrors instead of a gantry, which makes it super fast.

Specialty equipment definitely has its place. In fact, one thing we produce are higher end wood boxes for gift baskets and things. The slide top version is just nailed together, but we had a version that is V-grooved and miter folded. On the table saw it's a nightmare because the depth of cut is dependent on the material thickness and the material isn't consistent. The solution is for the cutter to be over the material, and I found an Italian company that offered an enormous machine with a vacuum bed to hold the material and 17 cutting heads, half of which applied in the X direction, the other half in the Y direction. It was actually an antique model they would refurbish, but each cutting head required its own dust collection which added up to a ton of CFM, and in our current space it was impractical all around. Then I found a company our of Canada that makes V-grooving machines that look like simplified CNC panel saws. It clamps the material down, and cuts across, then moves the material, clamps, and repeats. They could mount two heads so we could have a V-cut and a dado-cut. We would have bought that except the 3/8" Baltic Birch we would have used became unavailable and has remained unavailable to this day. Some day, it's still a direction we hope to go, but need the supply chain to catch up, and I saw the router as a stop-gap option to V-grooving. Though I watched a video where it did it with a saw in what appeared to be the most inefficient way possible. So I'm hoping there are better options. And maybe ultimately, a panel saw is a better option. There were some at IWF they showed making kerf-bending style cuts. It doesn't seem a stretch to mount a V-blade of the machine would support it. I do want to be able to V-groove acrylic as well. Boxes glued that way look so much nicer. We had constructed a machine to do it using a spindle mounted to a welded frame at 45 degrees with a 1/2" milling bit in it. It worked up to 1/4" thick material I think. But we were hand feeding the material and if we didn't hold it just right, the machine would grab the leading edge of the acrylic and tear it up. So we stopped using it, partly for safety, partly because it required cutting so many extra parts to get yield, it wasn't worth it, particularly when those were custom printed parts.

Basically, we want the most flexibility and capability we can get at this price point, while hoping to avoid any serious limitations that salespeople neglected to mention. From there, we will learn all it can do and push the limits like we always do. But it's valid to say that the process is complicated and time consuming because a lot of our projects are short timeline, and this is meant to help, not hinder those. In a simpler sense, the idea that 5-axis could do beveled edges, or diagonal drilling without extra steps, seems like a win. I'm not expecting 3D carving to be a simple process, but it should be a rewarding one.

My guess based on the breadth of your needs your not going to find one single best solution. The boxes likely a 3 axis with conventional folding via insert tooling, the 3D stuff I would still think a used Kuka style machine (especially if you could land the core machine for 10-20K) would be wise if you a have a demand for a realistic quantity of your three reference image type work.

Just my $0.02 based on a forum thread but I'd be cautious of 5 axis nested/pod rail unless your talking major production of a given group of items. Maybe you are and thats fine. Mind you you will need a draftsperson and operator along with that machine as with any other. The Kuka is not going to be suitable for miterfold box production even though Im sure it could be done.

I think Mark B is on the right track. To add to the argument for two machines; calculate the potential production time that would be run with sheet goods on a three axis. Then throw a 10-20 hour carve into the mix. Once you shift to a nested based production on a cnc it is hard to be without access to one in a workflow sense.
A worn-out Kuka from a production line or high precision environment is worthless to that sector but is a perfectly suitable machine for sculpting and can be had for very reasonable prices. Every video or photo I have seen of a robotic arm carving foam or wood is in a cage or walled off room to contain the dust and keep employees in one piece.

Tom, a 5 axis robot arm is far different than a 5 axis router.

https://www.youtube.com/watch?v=nNV9Zpjt7qs