John, thanks for that feedback. The complexity of the proprietary Camera Link interface and its specialized cables is often cited as a major reason why builders of machine vision systems look elsewhere for camera interface solutions. CoaXPress is certainly a majority contender to fill that need.
Another benefit not previously discussed is that CoaXPress can be used with simple, low-cost coaxial rotary joints. Camera Link, on the other hand, with its 26 conductors requires complex and costly slip-rings. This is of particular interest to the military, where the use of 360 degree rotating cameras is often required. Examples include gun turrets or for homeland security harbor surveillance.
William, that's a good question. It's not at all obvious, but "machine vision" usually refers to inspection during incoming materials/components, assembly, and manufacturing using computer vision technology. That said, the security area is another application for computer vision, but the requirements in terms of resolution and speed are much lower there than they are for product inspection. Even within product inspection, resolutions and speeds needed rfange widely, depending non whether the device under inspection (DUI) is a soda pop bottle or a semiconductor wafer.
The claimed speeds are very impressive, but the application in machine vision is a bit unclear to me. The immediate application that I can imagine is for recording vehicle crashes for vehicle safety development. That has been a big user of fast video recording for quite a while, and always looking for faster speeds. Of course, that area is less constrained by budgets, so it has not been the one to drive prices down. I suppose that really fast video would also be useful in other applications, but not always cost competitive.
Good questions, Alex and Beth. Since the standard was only finalized in March, and there are currently only a handful of companies making products, it's too soon to tell how fast the adoption rate will be or how widespread CoaXPress will become. It is expected to be very popular in geographic areas such as Asia, where there are a lot of analog cameras installed using coax cables. Promoters also expect it to be adopted by at least some first-time users. As in so many aspects of manufacturing systems, which standards are adopted, and what types of configurations are built, depend on many different factors.
Critiques of the protocol are also somewhat varied, depending on what it's being compared against. For example, promoters of both GigE/GigE Vision and Camera Link HS, the replacement for Camera Link, point out that CoaXPress, while serial, is not a point-to-multipoint protocol, something needed in the growing number of multi-camera systems to reduce cabling needs and speed data processing. Aside from those, additional critiques concern the protocol's SERDES topology, and start getting a bit esoteric.
How widely deployed is CoaXPress at this point among machine vision vendors, and what is the competitive spin/marketing angle that those vendors who aren't adopting it are using to push back against it?
Good questions, Beth. Yes, twice the per-cable speeds over cheaper cabling are some of the main benefits compared to the industry veteran, Camera Link. Camera Link requires its own special type of cabling which is pretty pricey compared to the ubiquitous coax. But there's another. CoaXPress cabling can carry that twice-as-fast signal 45 meters vs Camera Link's 10 meters before a repeater is needed.
So the benefit of the CoaXPress interface is better bandwidth and the ability to run over older cabling? For engineers designing machine vision systems, what kind of benefits does this bring to their applications--faster speeds at less reconfiguration expense? Any thing beyond that?
A new service lets engineers and orthopedic surgeons design and 3D print highly accurate, patient-specific, orthopedic medical implants made of metal -- without owning a 3D printer. Using free, downloadable software, users can import ASCII and binary .STL files, design the implant, and send an encrypted design file to a third-party manufacturer.
For industrial control applications, or even a simple assembly line, that machine can go almost 24/7 without a break. But what happens when the task is a little more complex? That’s where the “smart” machine would come in. The smart machine is one that has some simple (or complex in some cases) processing capability to be able to adapt to changing conditions. Such machines are suited for a host of applications, including automotive, aerospace, defense, medical, computers and electronics, telecommunications, consumer goods, and so on. This discussion will examine what’s possible with smart machines, and what tradeoffs need to be made to implement such a solution.