| Abstract This paper presents experiences gathered during realisation of the TEMPUS project S_JEP-09151-95 as well as former experiences of the author related to co-operation of institutions and individuals in the context of international research projects funded by European Union. |
Getting the TEMPUS Structural Joint European Project accepted.
Defining a proposal for a project funded by European Union, getting
it accepted, launching it and managing successfully is a difficult
task. Since 1995 my institute have been enjoying support from
European Training Foundation for realisation of the TEMPUS project
S_JEP09151-95. This project allowed me to fulfil my old
dream of organising at my institute a professional VLSI design
laboratory and introduce courses in this domain. Before presenting
the details of this project I would like to discuss certain issues
which I consider important for organising co-operation of several
institutions and for getting the project accepted.
First attempt
In 1992 the secretary of EUROCHIP program - Mr. Augustin Kaesser
encouraged our Institute to prepare a proposal for COST project
aimed at development of microelectronic education in Central Europe.
The proposal was prepared together with several other institutions
from Central Europe and with partners of EUROCHIP project, but
it was not accepted by European Commission in 1992 due to lack
of money.
Gaining European perspective
Later that year (1992) I was offered a job at University Joseph Fourier in Grenoble, France within a framework of European CAD Integration Project (ECIP - ESPRIT 2072). I grabbed this opportunity and spent a year in France. My responsibilities included logistics for VHDL standardization activities in Europe as well as dissemination of project results to European industry. I took part in preparation of the proposal for ECIP continuation (launched in 1994 as ESIP - ESPRIT 8370). I worked also on setting up an association called ECSI (European CAD Standardization Initiative), whose goal is to secure strong impact on world-wide standardization activities in the area of electronic design automation (EDA) for European companies.
My work for ECIP gave me opportunity to establish many personal links with people from different European companies and universities - those directly involved in ECIP activities as well as those interested in its results. During European Design Automation Conference in 1993 (EDAC'93) I got in touch for the first time with people from MASI laboratory (Université Pierre et Marie Curie, Paris) that were managing university booth at the exhibition that accompanied EDAC. I also met Prof. Rosenstiel from University of Tübingen there.
My work for ECIP gave me understanding of the process of preparation
of proposals for projects to be funded by European Union and procedures
of assessment of proposals as well as some knowledge about managing
international project. Thanks to my information dissemination
responsibilities I made a lot of professional contacts. And last
but not least working for ESPRIT gave me enough self-confidence
to think of preparing a project proposal by myself.
Second Attempt
Just after my return to Poland I participated in the First Workshop on Design Methodologies for Microelectronics organised in Gliwice and Cracow in October 1993. Prof. Rosenstiel and researchers from MASI were attending this event, too. Information about different possibilities of East-West co-operation had been presented during workshop by officials from Brussels and I was encouraged by Prof. Chojcan to apply for TEMPUS project. The proposal was created in co-operation with University of Tübingen and Université Pierre et Marie Curie (Paris VI) during December 1993 and January 1994. Most of the work I had to do personally and it was my fault that the final version of proposal was printed just a couple of hours after midnight on the 1st February, 1994. I missed the deadline - 31st January and the proposal was rejected for this formal reason.
At the same time however the money had been allocated by European
Union to support extension of EUROCHIP (now EUROPRACTICE) services
to Central and Eastern European countries. It was done within
the framework of Copernicus program (project CP 93 9093 - EUROEAST)
on the basis of our old proposal defined in 1992. The funding
for this project was approximately 10% of this original proposal.
Therefore no serious investments were possible, but we could at
least buy VHDL simulator and one workstation. We could also invite
a person from MASI laboratory to deliver a course on ASIC design.
This helped us to understand how we can benefit for co-operation
with Université Pierre et Marie Curie.
Copernicus proposal - gathering experiences
In the spring 1994 I took part in preparation of a proposal for
a Copernicus project. Several partners from universities and small
enterprises in western and central Europe were involved in proposal
preparation along with one partner from Russia. Our proposal was
rejected, but its preparation was a valuable experience anyway.
Successful TEMPUS proposal
The experiences I discussed above along with deeper analysis of
our needs and better understanding of the possible role of our
partners let us tune carefully our TEMPUS proposal for submission
in January 1995. The proposal was accepted several months later.
Our needs
The activities in the area of modern design methodologies for electronics were carried out at Silesian Technical University already in 80's. Hardware Description Language MODLAN was developed and partially implemented on our university mainframe computer by Dr Adam Pawlak as early as in 1982.
In the second half of the last decade a research project that addressed variety of issues related to electronic design automation of ASICs was realised at our Institute. It gave us good understanding of this domain, particularly in the area of VHDL modelling. However, it was impossible to organise real VLSI design environment with no access to UNIX machines, interactive graphic of reasonable quality and adequate development tools. We found PC XT not powerful enough for this purpose.
Therefore at the beginning of 90's we were still looking for external
financial support to buy necessary computing equipment as well
as for partners that could help us in organising education in
the area of ASIC design and high level design methodologies of
electronic systems.
Partner profiles
Meeting Prof. Rosenstiel and researchers from MASI laboratory
during workshop in Cracow was a key factor for choosing them as
partners in our TEPUS project. But the most important for us were
their experiences in teaching VLSI design and modern design methodologies.
The availability of free and proven academic software for supporting
these teaching activities was of course another important factor
to choose universities of Tübingen and Paris as partners,
as we were not sure of getting substantial funding. And last but
not least - THEY were willing to cooperate with us too.
Laboratoire MASI, Université Pierre et Marie Curie in Paris
VLSI design group of MASI Laboratory of Université Pierre et Marie Curie in Paris is managed by Prof. Alain Greiner. Over several years in the end of 80's and beginning 90's this group developed an Alliance design system and offered this system for free for other universities and institutions.
Alliance is a complete toolset supporting AISC design process
from chip specification in hardware description language down
to mask layout. It is based on a rigorous methodology that aims
at achieving IC designs correct-by-construction. Although Alliance
tools are not user friendly the supporting methodology is very
useful in teaching process.
Wilhelm-Schickard-Institut für Informatik, Eberhard Karls Universität Tübingen
Professor Wolfgang Rosenstiel's team is renown for knowledge and
experience in high-level design methodologies: behavioural synthesis,
hardware software co-design etc. We were offered for free Caddy
design system. Caddy is a high-level synthesis program developed
over last decade in Karlsruhe and Tübingen. Thus experiences
of our German partner complemented those of MASI laboratory.
Mobility program - advantages, problems, hopes
Important element of any international program is staff exchange
between cooperating institutions. In TEMPUS program mobility grants
are used to ensure accomplishment of transfer of teaching experiences,
to ease retraining of staff from CEE countries, to provide industrial
experiences for students and academic staff. European Training
Foundation considers this part of any project very important and
allocates a lot of money for mobility grants. Our experiences
show that realisation of mobility program faces many difficulties
in the context of TEMPUS project.
Mobility of staff from West to East drawbacks
It seems for us not easy to encourage people from western universities to visit Poland and stay here for longer period of time. In the context of TEMPUS projects most of such visits were meant to deliver training to our staff. Therefore the direct results of such visits bring benefit to institutions from CEE countries rather than to their guests.
That is probably why it is not so easy to organise these visits. Employees of western partner institutions have their own tight schedules, which do not leave much time to deliver courses for a CEE partner. Almost all of such visits to our University were shortened in respected to original plans due to different other obligations of scientists from our partner institutions.
I may only guess that these problems are not that serious in the
context of research programs when visits to CEE are meant to carry
out common research activities. In such a case the interest is
mutual.
Mobility of staff from East to West drawbacks
On the other hand I find possibility of retraining at partner institutions really advantageous. First of all they enable our staff members to gain access to rich research and teaching facilities of the partners, their libraries and experiences. It is much easier to ask for help foreign partners at their site than to make them come to Poland. Staying at partner institution enables to establish personal relations, which are very important for effective co-operation now and in the future.
I dare say that such retraining period abroad helps to break some psychological barriers by those that were never working before at any foreign academic institution. Usually it turns out that working together is easy and the existing technology gap between West and East does not mean any serious difference in skill or knowledge of the staff.
It is obvious that lack of necessary language skill makes co-operation
impossible. Thanks to visits to Western Europe staff members from
CEE countries improve their language skills. Last but not least
the retraining periods create opportunities to look at how the
research work and teaching is organised at partner institution.
The understanding of organisational issues seems to me very important
for any successful co-operation.
Mobility of students
To my mind it will take some time till exchange of students between Western and Eastern universities increases substantially. As there is still no agreed system of recognition of credits it is not obvious to send Polish students for studying at western European universities (and vice versa). An important factor that blocks exchange of students is a language barrier. It is even more serious in case of western students coming to Poland, although from this year on some courses at our faculty will be delivered in English.
It is much easier to organise student exchange for those who work
on Master theses and for Ph.D. students, but for the moment he
have no many experiences of this kind.
Practical placements
The industrial internships at western enterprises seems the most valuable experience for developing engineering skills of our staff and students. However, organising such practical placements proved to be nearly impossible. Partner institutions have their own difficulties in finding right practical placements for their students. Therefore they were not willing to help us in finding places for our students at enterprises in Germany and France.
We undertook this effort on our own. I must bitterly say that most of this effort proved useless. I sent out twice about 20 faxes to companies that specialise in ASIC design, but I did not get any single answer ! This experience shows how important the personal relations are. All companies that accepted our students and staff members: SIEMENS (Munich), VLSI Technology (Palaiseau and Sophia Antipolis) as well as Dolphin Integration (Grenoble) did so because I had some personal relation to people at these enterprises. What can be seen is that larger enterprises are more open to this kind of co-operation that SMEs. On the other hand they used to plan in advance, thus organisation of internship must be also planned at least several months before it starts.
What I consider a real success is that - after first experiences
- VLSI Technology and SIEMENS declared their interest in systematic
organisation of internships of our students at these companies.
Transfer of teaching technology
The most obvious benefit for us that resulted from our co-operation with Université Pierre et Marie Curie in Paris and with Eberhard Karls Universität in Tübingen was the possibility to use their teaching experiences in the area of our interest.
We got Alliance design system from MASI laboratory. We got trained in using Alliance in ASIC design and what is more important we got trained in underlying design methodology. The understanding of design methodology let us to define our own. We use Alliance with other tools that we have now at our disposal.
As we expected University of Tübingen were helpful in training us in high level design methodology. Their Caddy system of high level synthesis was installed in Gliwice and the intensive course on high level synthesis was delivered here. During my stay in Tübingen I took advantage of their experiences with Synopsys software, which we recently purchased.
We got from our partners slides and tutorials. They were useful
for preparation our teaching materials in Polish. It decreased
necessary effort considerably.
Achievements of TEMPUS S_JEP 0915 95
The amount of money allocated for our TEMPUS project was lower
than we applied for. Nevertheless more than 110,000 ECU was allocated
for buying equipment and software for three years. This is quite
large amount of money when compared to the investment budget of
our Institute and it enables us to establish and develop laboratory
of ASIC design.
Laboratory
We decided to base education on PC-based tools using both commercial software available through EUROPRACTICE as well as tools from our partners running under Linux system. However, for supporting advanced courses we chose professional, UNIX-based tools.
We bought more than dozen Pentium based PCs. All PCs have installed MS-Windows as well as Linux and may be used as X terminals. We purchased VHDL based simulation and synthesis tools that runs on PCs under MS Windows. This environment is gaining importance in EDA industry and in Poland it is by all means more common than UNIX. We take care to ensure smooth interaction of these tools with traditional schematic capture and PCB design tools as well as with tools for programmable logic design. We use also academic software from our partners under Linux operating system.
In addition to this we purchased several licenses of SYNOPSYS toolset to teach high level design methodologies and Compass toolset to teach ASIC design. We had two SPARC workstations before project started and we decided to buy one ultraSPARC server and another SPARC workstation (to be purchased this year) to run these tools.
This year we would like to buy tools for supporting analog IC
design, Multi Chip Module (MCM) design as well as microsystem
(MEMS) design.
New courses
The only course that provided students with information on modern design methodologies which was available before the start of our TEMPUS project was the course on Computer Aided Design of Digital Circuits. Its contents was modified over last two years. More information on ASIC design methodology and VHDL-based synthesis was introduced. Moreover - the course may be now much more effectively illustrated by laboratory exercises thanks to investments done.
This course will be available in English from the next year on. This may encourage foreign students to come studying at our University (within the framework of Erasmus project).
The course on Digital System Modelling introduces students extensively to modelling seen as important activity of the contemporary design process. Issues such as modelling for RTL and behavioural synthesis, modelling at gate level for sign-off simulation, using hardware description languages in testing, system level specification methods, modelling of mixed analog-digital circuits are presented during this course.
Elements of this course form a basis of the VHDL-based design course, which is prepared to be introduced into EUROPRACTICE portfolio. We believe that taking advantage of our modern laboratory we will be able to offer professional training to people from industry and academia in the area of using VHDL in design process.
The second course introduced this year is Introduction to ASIC Design. The syllabus of this course was defined within first year of the EUROEAST project but there was neither enough equipment nor software to offer it to students.
In the last year of our TEMPUS project we would like to develop
course(s) that would present to our students problems related
to analog IC design, advanced semiconductor and assembly technologies
like BiCMOS and GaAs technologies, MCMs and chip-on-board assembly
as well as microsystems (MEMS) design. Some of these subjects
go beyond experiences of our project partners, but we consider
them as important design technologies in modern electronics and
we want to introduce them to our students too.
Perspectives of long term co-operation with project partners
Our Institute will finish our TEMPUS project next year equipped with design technologies that enable us not only to teach effectively modern design techniques, but also to proceed with research activities related to VLSI design. I hope we will be able to use these capabilities along with established relations to continue and further extend co-operation with western universities. Contacts with our partners allowed us already to find several research topics of common interest.
The most interesting for me personally are activities in the area of high level design methodologies related to VHDL modelling and hardware software co-design. Different problems from this domain are addressed at all three institutions involved in our TEMPUS project.
The researchers from laboratory LIP2 (new name for MASI laboratory) started recently activities in analog IC design, which is a new domain for them. The analog design team from our Institute have vast design experience in the analog PCB design with standard parts. The research work on automatic synthesis of analog circuit structures done in our Institute complements research on automatic layout generation of analog circuits conducted in Paris. I hope that the co-operation in this domain between our institutions will start and will survive the end of our common TEMPUS project.
There is a strong research group in Tübingen that does research
on neural networks. I learnt about it during four weeks I spent
in Tübingen last summer. There is also a group of researchers
interested in this domain at our Institute. I am trying now to
bring both groups together and encourage them to start co-operation.
Conclusions
I am absolutely sure that developing relations between individuals is crucial for starting any co-operation between institutions. Over the first two years of our project realisation I learnt how important is the mutual exchange of staff for establishing links that could possibly result in long term co-operation and how difficult it is to implement.
Internet as a technology that offers easy and effective information access helps great deal in looking for partners. Nevertheless, personal contacts remain important in building up relations that may result in common research activities. Therefore the actions like those organised within the framework of BENEFIT project - I mean workshops and summer schools that bring together people with matching professional and research interests, are extremely important for establishing links between institutions and people from European Union and from Central and Eastern European (CEE) countries.
From my point of view Tempus and Copernicus projects proved to be very useful in modernisation of research and teaching technology that universities from CEE countries had at their disposal as well as in updating knowledge and experience of the their staff.
I do not know statistics that could tell us how effective these projects were in establishing links between universities and industry. I heard of model examples of Copernicus projects that allowed new technology transfer from university to SMEs in Poland. I am afraid that these examples are exceptions. I believe that Polish Committee of Scientific Research should encourage participation of Polish SMEs in European projects. Moreover, it seems also necessary to support a technology transfer to Polish industry. While presenting newly acquired design technology and capabilities to local electronic enterprises we face a huge technology gap in what we can offer now and what is needed and used in Polish industry.
I think we have a long way to go to overcome barriers in possible
co-operation between SMEs from European Union and academic institutions
from the CEE countries. As I pointed out above there was no interest
of SMEs in accepting our students for internships. We had however
good experiences in co-operation with industry, too. Development
of stable co-operation with industry requires still a lot of efforts.
BENEFIT
Special Report Main Page