Embedded Systems Engineering Standards Column vol 13.1 Jan-Feb 2005 A New year: New Beginings By Chris Hills

 

I have to give the usual disclaimer that that these are my own personal views and not those of the ESE Editor and publisher. or those of my employer.... see next months item!

 

After 6 years where I was I decided it was time to move on. I found somewhere [I thought was] up and coming to move to, resigned and spent December on gardening leave. Given that my idea of gardening is laying 3 cubic metres of green Readymix concrete and that it was December my gardening was relegated to looking at it though the window and planning summer barbecues! However, the sabbatical was very helpful. As I mentioned in a previous column "Two Weeks In The Summer" everyone needs a break from the stresses of modern life every now and again. Though an acquaintance took it a little more seriously and spent a year globetrotting with his family. He looks well on it and says it should be compusory!

 

STOP

 

(A word from my sponsors :-) Actually I put a lot of of the program together for the IEE one. This information for these came in too late for the printed edition. I will be at both these conferences. Chairing the MISRA-C Panel session at the IEE one and a colleague presenting at the other. Phaedrus Systems is strongly into high-integrity systems.

IEE Automotive Embedded Systems Conference 15Th March, Savoy Place, London UK http://conferences.iee.org/autoelec/

Medical SW Conference 17Th March Cambridge UK http://www.embedded-masterclass.co.uk

 

For this New Year I have new working conditions: I'm at home! I am fortunate enough to have a room dedicated as an office in the house. With cheap broadband, the low costs of networks and phones for many jobs you don't need to be in the same office. I actually have a better computing setup now than last year in a "proper" office. I have a printer, scanner, fileserver, fax and other tools I don't have to share with the rest of the company. I may only have a 512kb ADSL but I don't have to share that either so I have a faster overall connection. As I can redirect calls and email to my mobile I am never out of reach...Should I so choose.

 

I also have a Voice over IP system that I tend to have "open" to the person I would normally have a desk next to. Also we can conference call on it. So I still get all the office gossip. That said I find I have fewer distractions than working in a conventional office. For anyone with broadband and working remotely I can recommend VoIP phones

 

I also spend more time at home and more in the office. How? I used to travel an hour each way to work per day. I have cut this by 100% so I have an extra 30 minutes at home and get to the office 30 minutes early and no stress! The more relaxed start to the day I find gives me a less stressed day. It also has other advantages though on the down side the catering is strictly self-service! The other advantage is it is easier to deal with the out-of-hours calls to the Far East and the US… Nipping into the office only takes a few seconds. It is much easier when you have a problem you need to work late on. I can stop for dinner at home and go back to the office. A good way to get out of the washing up.

 

Others who work at home tell me, that discipline is a problem though I haven't found this. I do not have young children at home during the day and my wife works. Perhaps it is because I have a room that is separate as an office and 20 minutes of daytime TV is more than enough for a lifetime for anyone. Does anyone conscious or sober actually watch daytime TV? The other problem is that you can end up not coming out of the office. I know someone with his own business he could run from home but he rents a small office because he wants to separate the two lives

 

The only disadvantage to working from home is that my wife now expects dinner ready when she comes home from work! And the washing put through the washer and dryer. Though that is a small price to pay for not having to deal with rush hour traffic every day. It is also going to save me about 12,000 miles a year.

 

There are yet more New Year changes: On the Standards side I have finally stepped down as Convener of the BSI C panel after 4 years. Now I am just an ordinary C panel member again. I should have been convener for only three years but it is sometimes difficult to find suitable volunteers to take over. Running a standards panel can be like herding cats. You get an interesting mix: academics in ivory towers, academics in who fully understands the modern market driven higher education, industrialists who have to make it work (at minimum cost and time), those who produce compilers and tools that work to the standard and those for whom it is a hobby. For n experts on a language panel there are x+1 opinions. When you have academics from two universities or two tool vendors disagreeing there can be more than technical exactness involved. That is apart from the general politics of international standards panels.

 

We also have the problem of where C is going. There are a few people on the C panels who are also C++ people. Added to which the many of the C++ panel think that C should become a true subset of C++ by moving C to C++ rather than cutting C++ back. This will of course cause problems as it would take C a stage on from C99 at a time when all the worlds embedded compilers…. Actually I think it is virtually all the worlds compilers, embedded or not, are still only C90 (+ bits and pieces) compliant. If the C++ panel had its way the C that would result would not be used by the embedded, real-time or safety critical markets. In the same way ISO BASIC is not used. The world uses VB and a few non-ISO implementations of BASIC. So C, one of the world's most popular languages, would be practically speaking non-standard in its use. As a lot of the worlds safety critical software is written in C this is a bit worrying.

 

The UK C panel is currently undergoing some "readjustment" that is a little painful. There have been some disagreements and people not talking to each other As I said it's like herding cats. More on this next month when the dust has, I hope, settled and the blood washed from the walls.

 

On a brighter note another standards working group has been formed. It is the High-Integrity Study Group looking at the application of software in high integrity applications. The majority of these fall into the embedded and real-time field so I shall be keeping an eye on this. It will only be of use if it is intelligible and easy to implement for the average engineer working on critical systems.

 

There are many proven methods like Z, VDM and other Formal Methods that the mathematicians delight in because they can be mathematically proved. Though they never get widely used, let alone become a de-facto standard, because they are too abstract and theoretical. Also unless you are mathematically inclined it is easy to make mistakes and in any event they still have to be translated in to code at some point. MISRA-C1 on the other hand whilst not perfect (and didn't the experts argue) was easy for engineers to understand. We need a balance.

 

I have an email from someone involved in SPARC-ADA telling me that MISRA-C has probably done more to raise the standard of Software Engineering than SPARC ADA. Why? Because SPARC-ADA is used by say the top 5% of critical systems. It has improved them a couple of notches. Things at that level were carefully engineered anyway. However, MISRA-C has improved by, quite a few percent, the other 80% of projects written in C. So the over all effect is that MISRA-C has raised the lower and medium bar quite a bit whereas SPARC ADA raised the top few a little. It is a bit like saying SPARC-ADA gave the F1 teams ABS, Traction control and stability control. MISRA-C gave all road cars ABS. So I am hoping the new High Integrity Study Group will bring the average level up a fair bit rather than create another little used system that a very few will use to raise something that is near perfection just a little higher. It is all well and good getting the software in nuclear power stations from 99.4 to 99.6 but I would prefer to see something that gets automotive or marine SW from 50 to 80 on reliability. NOTE all figures illustrative. Please don't quote them!

 

I found an interesting comment on this in Safer C (Hatton 1994) which shows the fault rate for code produced at NASA Goddard from 1976 to 1990. Whilst the best programmers went from 6 to 4 errors per 1KloC the worst went from 12 to 5KloC over the same period. In a similar way I hope this High-Integrity Study Group will help bring the multitude up to the level of the best.

 

MISRA has now set up the MISRA-C Steering Committee as a legal entity to continue developing MISRA-C. It is separate from the original MISRA Steering Committee that produced the MISRA-Guidelines of which the C Guidelines were the 10th part. iSystem Ltd UK is one of the nine founder members of the MIRSRA-C Steering Committee. The other eight are the rest of the original MISRA Working Group and Dr Les Hatton who oversaw the first attempts at a MISRA-C test suite. See the web site http://www.misra-c.com for details.

In my last column I spoke of standards of education and the SEA program. I had a reply to tell me that there is not a shortage of Engineers. What do the rest of you think?

 

There may not be a shortage today but it is the next five years that worry me. James Dyson, in the Richard Dimbelbey Lecture on the BCC last month, underlined this in spades. The synopsis, if not the transcript at http://news.bbc.co.uk/1/hi/business/4081937.stm is well worth reading. His view is that we do not need to manufacture in the UK but to own the IP and have the R&D in the UK. His company, reluctantly, moved the manufacturing to Malaysia but the company in still UK based. Especially the Engineering and R&D. His secret? Investment in UK based IP and R&D. The average R&D spend in the developed world is about 8% of turnover. In the UK it is 4.5% at Dyson it is 12%. He had a plea for more and better education of Engineers in the UK. Not as it is now the UK higher education working on short term profit and loss training overseas students for cash. We need to enthuse the UK school kids that Engineering is a worthwhile profession to enter. So get in there and help promote Engineering to kids. By the way the IEE has just launched a magazine and web site (http://www.flipside.org.uk) for 11-14 year olds. Have a look. Give it to your kids.

China (again)… To add to my previous comments the feedback I am getting is that whilst the Chinese can manufacture cheaply they don't (yet) have the design skills. Several readers have told me about "cost reduction" teams in China that cost-reduce the systems to the point where they don't work reliably or in quite the way the were meant to or they use cheaper "similar" parts with a "slightly" different specification. They need technical expertise on the design stages that does not seem common in China at the present. This is probably why the Chinese are trying to buy Rover Cars. They will, like Dyson, design in the UK but manufacture in the Far East. The difference is the IP ownership will be in the Far East. The other difference is that China is embarking on a large-scale push in training Engineers in Universities and schools. A lot of them in UK universities.

 

We must if we are to survive, as I said last month, build up our support for engineering in education. Improve the status of Engineers as a career (see the column on licensing of Engineers). In the USA it may have happened already in some areas. An Item in EE Times for Asia

(http://www.eetasia.com/article_content.php3?member=no&article_id=8800354282&DD=1399dfb5)

on outsourcing had this to say "And we are starting to hear statements like the following: "I went to China because it is so much harder to recruit top talent in the U.S." The bottom line on outsourcing will not be cheaper labor or boiler-room operations; it will be that the United States has failed to train enough engineers for the next generation." At this point the work goes abroad because there aren't the engineers and as the work is going abroad no one wants to enter the profession it is a downward spiral

 

If the UK can wake up and train more engineers, invest in R&D we may just become the technological powerhouse to replace the manufacturing country we once were.

 

References.
Hatton, L. (1994). Safer C: Developing Software for High-Integrity and Safety Critical Systems, MCGraw-Hi

 

Author Details and contact

 

Eur Ing Chris Hills BSc CEng MIET MBCS MIEEE  FRGS   FRSA is a Technical Specialist and can be reached at This Contact

 

Copyright Chris A Hills  2003 -2008
The right of Chris A Hills to be identified as the author of this work has been asserted by him in accordance with the Copyright, Designs and Patents Act 1988