Sunday, September 14, 2008

What career paths exist in machine vision?

We can begin with the user. Typically this a manufacturing engineer who has some responsibility for maintaining, and possibly programming, a vision sensor or system. This guy just needs a grasp of the basics, so machine vision is something he gets involved in from time to time.

There is however a trend, especially in larger facilities, for a plant to employ a full time vision engineer. Someone in this role needs the skills to engineer a vision application from scratch, using either a PC platform or a range of ‘smart’ cameras. This means they may come from an electrical background, and may well have some plc and/or robot programming experience.

A significant risk in both these roles is the 3 a.m. phone call from the plant to tell you that the line’s stopped because the vision system is “failing everything.” If that kind of excitement is not for you, perhaps you need to be in the vendor side of the industry.

By vendors I mean both the distributors/resellers and the integration houses. At the distributor level the machine vision role is typically described as ‘Application Engineer’. The job here is to sell your employers’ goods and services by helping customers solve their problems. This demands a good grasp of the basics – especially lighting and optics – and broad product knowledge. What it doesn’t need much of is integration expertise, which I define as the ability to install the hardware and hook up all the signals so the damn thing works.

This leads us to the role of the integration engineer. This is the guy who does the hard work of actually delivering a working system to a customer. It can be tough work, demanding long hours, first class faultfinding skills, and a willingness to travel. But on the up side … well I’m sure there is an up side.

The remaining areas we should consider are in product development. To an extent this happens in universities, but I believe that most machine vision development work today is carried out by the manufacturers themselves. (By manufacturer I mean camera companies, ‘smart’ camera companies, and software houses.) While some of this work is related to applications, and so needs optics know-how, most of it is concerned with software and firmware. If you have a talent for math and want to develop new algorithms, or if the field of embedded software appeals, then this is where you want to be.

You might notice that I’ve left out the fields of self-employment and starting your own business. While machine vision offers many such opportunities, in part at least because the barriers to entry are low, I feel it’s not something a new entrant to our profession is going to consider. That said, these might be good subjects for future posts.

Well that’s it for our quick tour through machine vision careers. There’s a job to suit almost every technical aptitude, so the question we should consider next is, why go in to this field?

No comments: