Intralox belts are sold under all sorts of names...mecnasort and tabletronix were a few older ones I recall when I first started as an industrial machinist 15 years ago. The problem with intralox is theyre relatively sensative to weight and if one row or cluster of links goes down, it can very quickly take the entire table down.
Mecanums on the other hand are incredible. If you need to service a four thousand pound part, you can drag it around on a table of mecanums like its a childs toy. Mecanums pods are even individually addressable, so you can get an email or alert to exactly which one is under too much load, or failing.
These are so damn cool. My old FIRST team wanted to use these but the lack of traction was a real problem. Great for moving heavy & slow, not so great for fast turns (even with relatively low center of gravity).
That's awesome! I think the year we considered it was "Aim High" (https://en.wikipedia.org/wiki/Aim_High) or one of the adjacent years. We used our previous year's robot as a test bed and found regular wheels could too easily push our Mecanum wheel robot out of the way / knock it off balance. Do you remember which competition you used the wheels? What did you end up using for a driver control layout?
Same problem we had this year (Power Up). We built two drive bases, and when we tried the Mecanum one against our 2017 robot, it just got pushed right out of the way. No friction.
The field was carpet, and we put treads on the wheels for traction. Our most common problem was blowing our fuse due to using drill motors for the drive train! (Low speed, high torque, high power draw.) But until that fuse blew, those goals moved where we wanted them to!
It was our rookie year this year, and after watching how great some other teams were with mecanum wheels, we decided to look into them over the summer. To be honest, I'm just excited to try driving them :)
Yup, like the article notes, we use these on ships where there are often times when you need to transport large pallets of material through a maze of passages and you often don't have enough reserved deck space to accommodate the turning radius of a normal pallet jack or 'walkie'. Material Handling Equipment with Mecanum wheels are omnidirectional (no turning radius), so you can transit heavy/bulky cargo through very tight passageways that would otherwise be virtually inaccessible.
Neat! I built a little mobile robot that uses omni-wheels[1], a related style wheel. Build report here[2] (forgive the lousy website, I'm a mech guy just starting to learn software)
The wheels allowed some neat driving tricks, but it also made it really difficult to drive in a straight line. You don't get the normal tracking behaviour that wheels give you, and differences in traction from wheel to wheel also tend to distort your course. Adding a gyro for use in a closed-loop angular control system really helped this a lot.
Reminds me of Intralox belts[0], which are used extensively in manufacturing and warehousing. Intralox belts are conveyor belts made of tiny wheels pointed 90 degrees (usually) off of the direction the belt is traveling.
When reading the wikipedia page I was surprised to learn that: "In 1997, Airtrax Incorporated and several other companies each paid the US Navy $2,500 for rights to the technology, including old drawings of how the motors and controllers worked, to build an omnidirectional forklift truck"
This was more than 20 years ago, but $2,500 seems on the low end for a usage-license. This patent is pretty interesting after all. Does anyone have any knowledge in this area? Is that average for these types of patents, or is the Navy giving it away for cheap to incentivize new innovation?
'97 was 25 years after the original patent was filed, so the original patent was well expired.
I would guess that what they were buying from the Navy were literally just the implementation notes, because even if the patent covers all of the original, key, patent-able ideas, it can still be a bitch to go from that to an actual working implementation.
I see where they're going now that it's pointed out.
The X pattern defines a single point of control where the lines cross, while the 'diamond' pattern creates (roughly) four outer lines of contact and thus for a complex terrain beneath allows for better control in the location on top of that uneven field.
Thanks for pointing that out, might be unrelated but I wasn't impressed by the video because it didn't seem to work right, i.e. the robot didn't travel straight, just did a series of small turns to go in one direction.
If the original patent was November 1972, then surely it had long-since expired by 1997, which then begs the question of why it was necessary that: "In 1997, Airtrax Incorporated and several other companies each paid the US Navy $2,500 for rights to the technology"
It seems there is some complexity and other patents involved.
From their SB2 form:
A form of our Omni-Directional technology was originally patented in 1973 and was sold to the US Navy. We secured a transfer of this technology from the Navy in 1996 under the terms of a CRADA agreement (Cooperative Research and Development Agreement) and we have worked since that time to commercialize Omni-Directional products. We received 3 patents regarding the "redesign" of the wheel. In addition, we have a license agreements for the algorithms used to control vehicular movement, and a patent for this technology has been applied for. (...) The patent for the Omni-Directional wheel expired in 1990.
If I describe an API and the algorithm used to implement it (a patent) that doesn't necessarily mean the source code I used isn't still useful (a copyright).
They were paying for the Navy's implementation of the technology - a bunch of copyrighted notes, drawings, and details - rather than starting from scratch. They were not paying for the idea of the technology (because the patent had expired).
I'd expect it's a submarine, which are mentioned in the page you link, https://en.m.wikipedia.org/wiki/Submarine_patent -- largely only a thing in the USA's former first to invent system AFAIK.
They seem much closer to omniwheels, and largely inferior for the mecanum's actual purpose: if you check out youtube videos, mecanum-equipped heavy lifters (e.g. KUKAs on the A380's assembly line) are way more reactive, fast and flexible despite being humongous.
Yep you are correct. It seems like the tires would have to consist entirely of a fairly elastic material, preventing the use of banded or radial designs.
Neither surface nor weight seem to be issues: they're used for multisurface forklifts, they do fine on grass or dirt or sand, they're used for heavy lifters. They're expensive and heavy though.
But I'd be more worried about operating speed, mecanums have maybe 60~70% of the traction of a regular threaded wheel, and probably behave much more poorly at high speed.
On a regular car, you're probably better off with auto-parking (the tech is getting pretty damn reliable, it's just a question of having it trickle down from the luxury segments) or even just parking assist.
Mecanums on the other hand are incredible. If you need to service a four thousand pound part, you can drag it around on a table of mecanums like its a childs toy. Mecanums pods are even individually addressable, so you can get an email or alert to exactly which one is under too much load, or failing.