Beginning of this week, I was together with two of my colleagues at the kick-off of the universAAL project, which is an integrated project in the field of ambient assisted living. It aims at consolidating a common infrastructure for AAL systems based on the results of the FP6 projects SOPRANO, PERSONA, mPOWER, GENESYS, among others. The meeting was very well organized, and the motivation and commitment of the team (among them SINTEF, Fraunhofer IGD, ITACA, Ericsson Nikola Tesla, IBM Research Haifa, ProSyst Software) was very promising. We bring in our recently lauched openAAL platform, an open source middleware for ambient assisted living.

The kick-off took place in Røros, a small town in the mountains of Norway, which is a World Cultural Heritage because it has been shaped by its several hundred years of copper mining. It is a beautiful location, but for most of us this has also represented a new personal record in terms of temperature. We had a small tour to the “smelte” in the evening at –34°C, and even during the day, the temperatures were around -24°C.

Last year, we opened our living lab for ambient assisted living which we have built based on our openAAL infrastructure.

Here is now the video on YouTube:

Last week I was at the SOPRANO Conference, which presents results of our AAL Integrated Project SOPRANO to the interested public. It was colocated with the fair on home automation Beurs Domotica & Slim Wonen at Eindhoven. Apart from the key SOPRANO contributions, we had presentations from the European Commission on their strategy of facilitating the development and deployment of AAL solutions (Peter Wintlev- Jensen), on general acceptance issues by Heidrun Mollenkopf (BAGSO) on ethical aspects of AAL research (by Marjo Rauhala from Vienna), and from the PERSONA project. It really has given a good overview of the current state of research, and the open issues. Particularly the lack of deployment was addressed as part of the panel discussion. There were several opinions on this:

• Lack of societal awareness about the problems that demand for AAL solutions (we will not be able to deliver care in the same way as today as there will be many more older people than now). This leads to a lack of political support.
• Lack of awareness by the immediate carers about the possibilities, availability, and cost/benefits of AAL solutions and potentially threatening change of the role of the carer
• Too high a pace of innovation, which leads to hesistant investments
• Immaturity of the technology as to practical usability and cost effectiveness
• Lack of direct end customer approach (start with the more prosperous as early adopters)

Probably it is a mixture of all of it, but I found particularly the first one convincing as it changes the perspective: we now longer ask what added-value this technology can deliver in addition to care services and informal carers, but we ask how can we keep up a similar level of care for a much larger number of older people with fewer younger carers.

While MATURE has always been inspired by bottom-up developments (and the concept of knowledge maturing has this an inherent assumption), we have always emphasized the importance of top-down activities as well. We have avoided to use the term “management” here, but rather used the term guidance for that. So far, we have mainly concentrated on value-based and valuation-based guidance (showing by appreciation what is considered good practice, which is usually subsumed in a notion of team/corporate culture), and structural guidance (i.e., the establishment and nurturing of communication structures). Furthermore, we have been struggling with potentials and dangers of incentive structures, mainly monetary/material and career incentives.

This week we were at the Professional Training Facts 2009 in Stuttgart (see here for a summary of MATURE activities at this event). This was a great opportunity to think and discuss about topics around competence development in company. One trend I have spotted was the increasing importance of indicators for competence development and the incorporation of those indicators into management instruments like management-by-objectives. At first sight, this always seems to be a good idea to “professionalize” the learning element in a company by making it measurable. This originates in the assumption that “you cannot manage what you cannot measure”, which is probably true when you want to manage things. The approach promises transparency and can be a step towards calculating an ROI for learning. The big problem, however, does not lie in the approach of defining indicators and measuring them, but rather in the concrete indicators themselves. These indicators do not naturally naturally derive from the topic at hand, but are actually always bound to a notion of an ideal state; they contain the statement: you should have a high score in this indicator – this would be the best option. This is not bad as such, but this fact is rarely reflected, especially because this ideal state is actually context-dependent. It can be different for large vs. small organizations, for innovation-focused vs. efficiency-focused organizations, for service vs. production etc. What happens is that somebody defines (probably with good reasons) a certain set of indicators, and other simply take those indicators and apply them without questioning their value for their context.

What does this have to do with knowledge and knowledge maturing? This has three aspects:

• This “ideal state” conception as such is a body of knowledge, and it has to be carefully examine if the underlying knowledge about the ideal state is has reached a level of maturity that allows for a standardization in which you usually simply take things and apply them (like, e.g., for many financial indicators). Or if we are on a lower level of maturity and have to develop from there our own answer to the question “what is the ideal state”. In this learning and maturing process we have to learn about the contextual factors that differentiate us from others. And even if we take and apply standardized things, we should allow and encourage questioning usefulness at any time.
• Indicators are not only about measuring, they are about management and guidance. They aim at changing the behaviour by explicitly or implictly encouraging to become “better”. Even if we know sufficiently about the ideal state, do we know enough how a certain set of indicators (potentially tied together with a complex formula) influences the behaviour? Is our knowledge about that mature enough to make them the basis for formalized instruments (like reward schemes, but also career decisions)? Can we differentiate between correlations and causal relationships? Can we separate external factors? Or should be modest enough to consider them what they are: indicators that measure something, but not the wealth of reality, and use them as a reflection instrument – and in the end maybe come to the conclusion that they measure nothing of interest.
• We are currently researching indicators for knowledge maturing both in the empirical and the technical-conceptual strand of the MATURE project. We should be aware that indicators always derive from a concept of ideal state, which is difficult to envision as a whole. So we will base those indicators based on our pre-conceptions (which has a lot to do with our value systems) – and we should carefully reflect on this problem.

As a conclusion: measuring can be very helpful for many aspects, also on the soft side, but we should understanding the development and application of such measuring instruments as a collaborative learning process which should involve many. Then this process and its result can be also a good guidance instrument.

Today I was invited to a meeting of the IEEE LTSC WG20 for starting a liaison activity between the various stakeholders in competency-based data management. Its goal was to form a group for developing a shared conceptual model. I have presented a very brief summary of our work in that area:

Here is also the position paper.

One of my main points was that we need to look at the use cases and their specific requirements as many of the misunderstandings come from the implicit assumption of a certain use case. Our initial analysis of use cases is reported here:

Simone Braun, Christine Kunzmann, Andreas Schmidt
People Tagging & Ontology Maturing: Towards Collaborative Competence Management
In: David Randall and Pascal Salembier (eds.): From CSCW to Web2.0: European Developments in Collaborative Design Selected Papers from COOP08, Computer Supported Cooperative Work vol. , Springer, 2010

Some use cases, like people finding, only rely on very weak notions of interests, while others – like career planning and rewarding schemes – rely on sound competency definitions. This is very important to understand – because all of them tend to speak of competencies. This also helps to understand why in some cases >700 competencies are appropriate, and in others 20 might be sufficient.

For a general conceptual model, I have pointed out the following challenges:

• Competencies are cultural abstractions
• Competency definitions are implicitly contextualized and a certain degree of ambiguity will always remain.
• Competency definitions are purpose-driven conceptualizations
• Competencies are time-dependent conceptualizations

Finally, I have managed to complete the last step of my PhD (my defense was already in February): I have published the work. My thesis is titled “Situation-aware information services for work-integrated learning” (supervisors Prof. Peter Lockemann and Prof. Johannes Magenheim) and is available in German. It summarizes my research on context awareness and context management, and on supporting work-integrated learning (e.g., the knowledge maturing model, but also the methodology for context-steered learning) with competence ontologies. Yes, it has taken much too long, even the step from 98% to 100% has taken me one and a half years. But you should be aware that either you complete your PhD quickly, or you take care of a research department and several interesting projects – and do many other interesting things…

I am in the lucky situation that this is just the starting point for my further work. Within the MATURE project, I can continue my research on knowledge maturing and on competency-oriented approaches. Additionally, it has turned out that the context management approach is ideal also for the domain of ambient-assisted living (AAL), which has been shown in SOPRANO, and will continue in UniversAAL. Context-aware system behaviour in context of adaptive user interfaces has been the subject in AGENT-DYSL and will continue to be explored in the upcoming myUI.

Currently I am travelling to Brussels for the negotiations of two successful EU proposals in the area of eInclusion – with a 100% success rate Call 4 of FP7 was a very efficient call Both proposals provide the opportunity to explore new paths in my second main stream of research: context-awareness. While I have started in the domain of technology-enhanced learning, the ambient-assisted living project SOPRANO has shown that the results (particularly the blackboard-based approach to context management, which allows for a combination of ontology-based techniques with statistical approaches and provides native support for uncertainty and the temporal dimension) can be easily transferred to ambient technologies. Also adaptive user interfaces are in need of a flexible context management system, as our first attempt in AGENT-DYSL has shown, which was aiming at adaptive reading support for children with dyslexia.

• In the upcoming AAL project universAAL (an Integrating Project – IP) we aim to develop our SOPRANO Ambient middleware (soon to be release as openAAL – an open source middleware for ambient assisted living) to become part of a reference architecture and open source implementation of a universal AAL infrastructure, together with a promising consortium of 18 partners.
• Within myUI (a STREP), we aim at “synergistic user modeling”, i.e., device-independent capturing of the user’s context so that we can more easily engineer adaptive user interfaces for various devices, particularly for users with special needs, like the elderly, but also others.

I am already looking forward to those projects, which will probably start in the first quarter 2010, although this will again mean an increased number of travels (which I could successfully reduce in 2009).

Last week we were at Graz, first for a MATURE Consortium Meeeting and then for the I-KNOW conference, which I always enjoy for its atmosphere. It is far more relaxed and suitable for networking with long lunch and coffee breaks in the afternoon. Unfortunately, the quality of the talks did not live up to my expectations  based on previous years’ experience (despite the fact that the MATURE project contributed 7 presentations and one poster presentation). This is strikingly similar to the WM 2009 in Solothurn. Is this a (rather alarming) indicator that traditional knowledge management forums do not attract the top research contributions? Or is the topic as such no longer fashionable?

On the other side, the event hosted the kick-off event for the Special Interest Group on Professional Learning (www.sig-protel.eu), which tries to increase the visibility of the topic on a European level, first by better networking among the concerned European research projects like MATURE, APOSDLE, ROLE, and others. In the discussion, it has turned out that despite the ambiguity of the term, professional learning seems to be umbrella term for KM and workplace learning. This SIG is a promising sign for a maturing community.

This year, Christine and I were giving a talk on integrating motivational aspects into the design of informal learning support, which reported on our findings on how to integrating motivational measures into tools for informal learning (the paper is available from here). Christine has done most of the work in ethnographic studies and their analysis. Currently, together with our colleague Athanasios, they are struggling to integrate their ideas into the four demonstrators of MATURE Year 2 demonstrators.

It has been an exciting year, working as the scientific coordinator and (as Pablo Franzolini, the other part of the management team, likes to put it) the “moral authority” of the MATURE project. The team consists of more than 50 people from various background plus an additional group of associate partners of more than 30 organizations. This team is actually the best team I have ever experienced in a research project: highly motivated, creative, with renowned specialists in their field – and a team spirit. The coordination job does not become easy, but your effort becomes really rewarding. After 12 months, it is now time to reflect on it:

Openness and participation vs. managing complexity. For MATURE, we have chosen from the beginning a very open collaboration style with other initiatives. Mainly through our associate partner network, we have established contacts with various companies, research institutions and other initiatives. This gives you the opportunity to discuss your ideas with many different people and collect various constraints and requirements from their practical experience. You can also start joint activities that yield mutual benefit to increase impact and quality. However, such an approach is always at the edge regarding complexity. An Integrating Project with around 50 active participants is already a complex environment, but the open approach that complexity with different timelines, different goal structures etc., which have to be aligned.

Guidance according to the seeding-reseeding-evolutionary growth model. What is the role of a coordinator in such a project with high degrees of uncertainty? In “The Effective Executive” Drucker puts forward the main guiding question: “What contribution from me do you require to make your contribution?” So, a coordinator needs to act as a moderator of parallel strands of activities, giving guidance only where the lack of guidance would prevent people from making their contributions. Fischer’s SER model, which we have used for MATURE, seems to be a good model. It consists of three recurring phases in creative processes: seeding refers to given a group an initial starting point to start working on, evolutionary growth is the phase where the group develops their ideas without external intervention, and reseeding is an intervention in which the results of evolutionary growth are examined, pruned, and new input is provided to the group. We have used that paradigm both for the ethnographic study analysis, and especially for the use case development. In the latter case, groups of application and technical partners were asked to come up with use cases for technology support that improves knowledge maturing. The basis (=seeding) were the personas from the ethnographic studies and a framework for use case descriptions. After a phase of evolutionary growth, use cases were examined, clustered, merged and a refined description framework was given. The groups then worked on those use cases again. This has yielded very rich use cases and negotiated use cases (technically feasible, challenging from a research perspective, and based on needs of application partners).

Role of collective sensemaking. In previous projects, I have already learned how important collective sensemaking is for a project’s coherence and building a real team. Sensemaking here mainly refers to making sense (in the large) out of what we do (or have done) in smaller-scale situations. For the degrees of freedom and the level of uncertainty in research projects, top-down planning (the “master plan”) is simply not appropriate:

• Top-down planning usually provokes the “following a plan” behavior, taking care of your own activities only, rather than caring for the project as a whole. People usually feel that something is wrong with the master plan, but are not subversive enough to challenge it.
• Top-down planning does not adequately reflect the learning process within the project team. Plan revisions are usually slow and cumbersome so that you rather stick to the old plan.
• Top-down planning avoids team discussions and decisions what is right or best by choosing something mediocre. Following a plan is way easier than having a planning framework within which you can and have to do your own planning in cooperation with others that are related to your activities.

A far better instrument than planning is the synchronization and coordination of activities through collective stories that bring together the individual pieces into a coherent whole. Such story building and collective sensemaking is a continuous process that is sufficiently agile to incorporate new insights and environmental changes, but ensures project coherence.

Last week we had the very successful second annual review of the Integrated Project SOPRANO in the field of Ambient Assisted Living. Our approach to a semantic middleware based on OSGi has been highly appreciated. At the core is on the one side a context management approach, which is largely based on my PhD work and has been developed further by my colleague Peter Wolf, on the other side the semantic service description approach from DIANE, which Michael Klein from CAS has developed. Our lightweight semantics approach on both sides plays very well together, as the following presentation shows:

We want plan to make this technology available to everyone. Stay tuned.