An interesting post Andy, particularly the bit about the BBC and SSL. Strangely I too worked there, possibly a bit before you, or you have met me over a 9K. You obviously think I am in some way out of touch with "the modern way" but this is far from the truth. Whilst one can simulate circuits very well with PSpice, it is not going to get one to design something completely new. It does not suggest where to start, or how to for example encode video at a low bit rate. I worked on a very difficult problem at the time, how to make a video rate A to D converter with at least 8 bits of accuracy and very good signal to noise ratio. I worked on audio converters with more than 16 bits, and excellent dynamic performance. The problem is not the circuit design, it is how to make components that are not fast enough or not accurate enough do something which they cannot do on first inspection. That is the engineers' job. I worked on chip design for Oxford Semiconductor, same again, we could simulate the silicon performance fairly well but the techniques to improve the performance require fundamental knowledge of things that are not built into the tools.


At first inspection pretty much anyone can do some electronic design nowadays, and probably make something which works. This is essentially the routine boring part which one never enjoyed, certainly I didn't. The interesting part is the innovation, and this is the part that is probably born, or at least developed at a young age. In some ways new engineers are overwhelmed by capabilities, you get large FPGAs, analogue components with fairly amazing specifications, processors with a few billion operations per second, as much memory as one can fill, and all for a very low price. What could go wrong? In fact everything is wrong. Take smart meters, why is the original specification so badly flawed that the products have to be sold with promises they cannot achieve? Why is wi-fi not working very well for many people? Why is the software delivered to me obviously faulty the moment I use it, the same with web sites, enter something unexpected somewhere and they go very strange indeed. I entered a postcode for address lookup into one yesterday with lower case letters, it said that postcode didn't exist! Is it so difficult to avoid that problem because it doesn't even matter to the post office! I normally estimate that it will take ten times as long to test a piece of software as to write it. So something which takes a bit over a week to write will take something like 3 MONTHS to properly test. The same for a chip design, test is a long and painful job because one MUST find the unexpected. On the 9K it took many months to make the automation work properly, and it wasn't finished when I left.


The thing that is important, particularly if you want to get more women into engineering is to concentrate on the innovation part and not anything else. Anyone (men or women) will already basically be interested in engineering anyway, this is not a career choice to be made lightly, because it is not well paid in general, is not suitable for leave due to children. A few years leave will make someone relatively useless as you know, due to the extreme rate of change of electronics and many other areas of engineering. Some areas may cope but not the leading edge of technology.


Engineering has never been easy, Brunel and Watt took enormous risks and killed themselves to make progress. It is the same now, engineering is a hard career, and you will notice that those who cannot do it rapidly transfer to "management", usually being very bad at it because they do not have the understanding of the discipline. You may disagree, but as proof I hold up the IET attempt to entice women, some are very good, but the number is low and there is not much interest.

An interesting post Andy, particularly the bit about the BBC and SSL. Strangely I too worked there, possibly a bit before you, or you have met me over a 9K. You obviously think I am in some way out of touch with "the modern way" but this is far from the truth. Whilst one can simulate circuits very well with PSpice, it is not going to get one to design something completely new. It does not suggest where to start, or how to for example encode video at a low bit rate. I worked on a very difficult problem at the time, how to make a video rate A to D converter with at least 8 bits of accuracy and very good signal to noise ratio. I worked on audio converters with more than 16 bits, and excellent dynamic performance. The problem is not the circuit design, it is how to make components that are not fast enough or not accurate enough do something which they cannot do on first inspection. That is the engineers' job. I worked on chip design for Oxford Semiconductor, same again, we could simulate the silicon performance fairly well but the techniques to improve the performance require fundamental knowledge of things that are not built into the tools.


At first inspection pretty much anyone can do some electronic design nowadays, and probably make something which works. This is essentially the routine boring part which one never enjoyed, certainly I didn't. The interesting part is the innovation, and this is the part that is probably born, or at least developed at a young age. In some ways new engineers are overwhelmed by capabilities, you get large FPGAs, analogue components with fairly amazing specifications, processors with a few billion operations per second, as much memory as one can fill, and all for a very low price. What could go wrong? In fact everything is wrong. Take smart meters, why is the original specification so badly flawed that the products have to be sold with promises they cannot achieve? Why is wi-fi not working very well for many people? Why is the software delivered to me obviously faulty the moment I use it, the same with web sites, enter something unexpected somewhere and they go very strange indeed. I entered a postcode for address lookup into one yesterday with lower case letters, it said that postcode didn't exist! Is it so difficult to avoid that problem because it doesn't even matter to the post office! I normally estimate that it will take ten times as long to test a piece of software as to write it. So something which takes a bit over a week to write will take something like 3 MONTHS to properly test. The same for a chip design, test is a long and painful job because one MUST find the unexpected. On the 9K it took many months to make the automation work properly, and it wasn't finished when I left.


The thing that is important, particularly if you want to get more women into engineering is to concentrate on the innovation part and not anything else. Anyone (men or women) will already basically be interested in engineering anyway, this is not a career choice to be made lightly, because it is not well paid in general, is not suitable for leave due to children. A few years leave will make someone relatively useless as you know, due to the extreme rate of change of electronics and many other areas of engineering. Some areas may cope but not the leading edge of technology.


Engineering has never been easy, Brunel and Watt took enormous risks and killed themselves to make progress. It is the same now, engineering is a hard career, and you will notice that those who cannot do it rapidly transfer to "management", usually being very bad at it because they do not have the understanding of the discipline. You may disagree, but as proof I hold up the IET attempt to entice women, some are very good, but the number is low and there is not much interest.