It is sometimes a source of wonder how acronyms are created. ARTEMIS, the European programme to promote embedded systems, was formed from Advanced Research and Technology for Embedded Intelligence and Systems. Which came first the chicken or the egg?
The ARTEMIS programme’s roots are from around 2004, when the European Union recognised that embedded systems were a vital part of the economy, and, while there is a great deal of expertise in Europe, it is scattered across different organisations – commercial, governmental, and educational – and across multiple countries. The underlying idea was that by co-ordinating and supporting research work, European industry would be better prepared to compete in world markets.
Let’s start with some terms of reference for non-European readers. The EU/European Union/Common Market is, according to its web site, “… a unique economic and political partnership between 27 European countries that together cover much of the continent.” (At least that is the English language version – the web site exists in 22 other languages, so the brief definition could equally be, “Comhpháirtíocht uathúil gheilleagrach is pholaitiúil idir 27 tír Eorpach, ina bhfuil formhór na mór-roinne, is ea an AE.” Or “ES ir v?l nebijusi ekonomiska un politiska partner?ba starp 27 Eiropas valst?m, kas kop? aiz?em liel?ko kontinenta da?u.”) The list of things that the EU is involved in goes on for pages and pages and, depending on where you stand politically, it is either generally a good thing or an economic albatross around the neck of the European taxpayer, interfering in local laws and siphoning cash from hard-pressed tax payers in my country to wastrels in your country. My personal feeling is that an organisation that has helped to stop European countries from fighting one another for nearly 70 years has to be a good thing. (Although, like many large organisations, it could do with a bit of a shake-up.)
The core of the EU is the single market: the (mostly) free movement of people, goods, and services between member states. This simple concept has a number of ramifications; for example, harmonisation (a good EU word) of standards for safety. It also extends to co-operation to help EU remain (or become) competitive in different markets. And this brings us back to ARTEMIS.
A few weeks ago, ARTEMIS held a review event in Paris, where demonstrations of completed and on-going projects showed what they had been working on. There were presentations on different aspects of the programme, and the new Strategic Research Agenda (SRA 2011) was discussed in depth. (One of the problems with ARTEMIS is that it has developed its own jargon. This is a subset, with its own extensions, of the huge jargon built up by the EU. I will try to avoid using the jargon where I can, but there are times when there is no choice.) The SRA is the framework drawn up by the ARTEMIS Industry Association members – the over 200 companies and organisations that have joined the ARTEMIS IA. And against this, the funding teams will assess proposals for research projects and programmes.
The original SRA was drawn up in 2006, and since then, things have moved on. In particular, interconnectivity and the Internet of Things have become more important, and SRA 2011 talks about embedded systems as “becoming the neural system of society.” And one of the key things that I learned was that ARTEMIS is not just about technology; it is also about technology contributing to society. This is twofold: providing Europe with an income from being competitive in a worldwide market and developing systems that help European society generally. These last are listed in SRA 2011 as systems that assist in health and well-being, providing transportation that is green and supportive, and creating the smart buildings and communities of the future.
ARTEMIS research projects are meant to be “Pre-competitive” – that is, to provide a common basis of knowledge that commercial organisations can use to make real products, although there are suggestions that this limitation may be loosened.
There is a formal process for preparing projects and applying for funding within the terms of the SRA. Early in the year, there is a “Brokerage Event” (I warned you about the jargon). This is, in effect, a dating session where organisations hitch-up to prepare proposals. In the UK, there is a pre-briefing event designed to encourage British organisations to get involved in advance of the Brokerage Event, and other countries run similar sessions. Proposals, both those resulting from the Brokerage Event and other activities between organisations, are sent to ARTEMIS in response to a “CALL,” and finally, successful proposals are funded, partly from the central EU funds, partly from local national funds and, of course, by the organisations involved committing time and effort. The process is underway for CALL 2013, and the Brokerage Event is in January, at a hotel at Heathrow Airport, London. In addition to the guidelines of SRA 11, there is a bias towards funding projects that are building on the results of previous projects.
So far 44 projects have been funded. Some, particularly from CALL 2008, have been completed, and others are still very much underway. There is a wide spread of different activities, from robotics to remote health monitoring; from tools and tool-chains to multi-processor implementation.
Two ARTEMIS projects, in particular, caught my attention, CESAR and VeTeSS.
CESAR (Cost-Efficient methods and processes for SAfety Relevant embedded systems – see what I mean about acronyms?) is now finished. It had a budget of 16.1 million Euros (US $21 million at today’s exchange rate) and involved 55 organisations across ten countries. Organisations ranging from the big chip company Infineon, through large manufacturing companies (such as Airbus and Thales), tools developers (such as Esterel), to small companies and universities. The argument for CESAR was that the safety-critical market has a proliferation of tools, processes and methodologies, which are not integrated nor are they interoperable. CESAR aimed to create a Reference Technology Platform (RTP) that would “embody meta-models, methods, and tools for safety-critical hard-real-time system development supported by European tool vendors.” This RTP would exist initially for domains such as aerospace, automotive, and rail transport, and it would extend later into other areas. There was a strong emphasis on requirements engineering, model-based development, and components-based developments. Looking on their website, there are a lot of publications listed that should demonstrate how CESAR has met these objectives, but many are marked as restricted – that is, available only to those who had taken part in the programs. Some interesting ones that are publicly available are surveys of the state of the art in, for example,Safety and Diagnosability and Modelling Languages, but these are now over three years old. I don’t remember being made aware of them when they were published; otherwise I would have written about them.
CESAR has now finished, but there is a sort of follow-on in VeTeSS (Verification and Testing to Support Functional Safety Standards). This is an ongoing project “…to develop standardized tools and methods for the verification of safety properties, particularly for generic components and subsystems used in safety-relevant embedded systems.” This a much smaller project, with eight countries and 24 organisations, again involving Infineon and this time NXP, plus other companies, including several automotive companies, and several universities. This is due to run for 36 months, having started in May 2012. The research areas are listed as:
Common Cause Failures
Fault Coverage Models
Executable Safety Case
Making ISO 26262 Usable
The project will work as a matrix with Technical Streams, looking at methods, and Work Packages, looking at software, hardware, and system testing. Interestingly, one of the technical streams will be looking at formal methods. I will be monitoring this project, if I can, and I hope to bring you more news in due course. And I have also heard informally about another, larger project that should be announced soon, in the area of safety-critical systems development.
Concerns that I have heard expressed about ARTEMIS are that contributing organisations find that they have to put in a lot of effort that does not contribute directly to the project: people point to the hoops to jump through to pull together a consortium and write a proposal, the issues of co-ordinating the work of many different countries with different languages and cultures, and the reporting and auditing requirements of ARTEMIS. (This last is something that the ARTEMIS management team has said it is aware of, and it is looking at ways to apply a lighter touch while still making sure that projects are on course and spending money wisely.)
The cultural differences were discussed at a small meeting, and there the story was told of three senior ARTEMIS people in Amsterdam about to cross a road when the pedestrian light went red. Immediately the Frenchman was on the other side of the street, the Dutchman was on the central refuge and the German (who told the story) was standing stolidly waiting for the light to change back to green. More seriously, engineering culture does provide some common ground, and one of the benefits of the projects is that those involved get a wider perspective, which helps to improve their ability to compete in world markets.
The bottom line is this: Is the investment of many hundreds of millions of Euros worthwhile? Many of the companies involved certainly think so, particularly some of the larger and multinational companies, and they are willing to put up with the overheads for the benefits they get. As a European taxpayer – what do I think? On the whole – yes. But it is clear that some governments are more energetic at encouraging companies to get involved than others. And also, the results of the work are not widely disseminated. There are wonderful glossy publications, and many conferences have papers from members of ARTEMIS projects. But it is not hard to escape a feeling that the members of ARTEMIS are talking mainly to each other and to other European projects, like ITEA. (I quote: “…a strategic pan-European programme for advanced pre-competitive R&D in Software-intensive Systems and Services (SiSS).”) As I noted above, there is a lot of material from the completed programs that is restricted. So the one big improvement I can see for ARTEMIS is greatly improved communication across the board. But then, I would say that, wouldn’t I?