As I wrote in the posting Microblogging – What’s next?, I strongly believe that there could be huge value in including non-human information sources to (enterprise) microblogging. The original idea for such a scenario including processes, machines and sensors (see the image for examples from Twitter) reaches back to my master thesis one year ago. It took that time that I finally came up with a term which could name this idea: ubiquitous microblogging.

Obvious, ubiquitous microblogging leans on the well-known research field of ubiquitous computing. While the latter understands ubiquity in a way that artificial computing devices are everywhere in the real world, the meaning of ubiquitous microblogging is that of real world objects being integrated and represented in an artificial computing space. In our definition, ubiquitous microblogging means a microblogging system including everyone and everything in an organisation. Therefore we borrow the conceptual meaning of ‘ubiquitous’ in the sense of its Latin origin ‘everywhere’.

Weiser (1991) in his vision about ubiquitous computing stated that ‘the most profound technologies are those that disappear’. Figuratively speaking, this is also true for ubiquitous microblogging as the goal behind our approach is to hide real world’s information access complexity with providing a flat information space accessible by the easy following-mechanism.

The approach of ubiquitous microblogging has much to do with the search for enterprise use cases of microblogging and a rising number of researchers is thinking about this topic. Michael Rosemann from Queensland University of Technology described how microblogging could be used for business process management. Alexander Dreiling from SAP shows a prototype for collaborative modelling with Google Wave (is Wave microblogging? I am going to discuss this question in a future posting). But the other way round is also possible, as the guys from Akibot show with their microblogging bot using NLP (Natural Language Processing). And finally, our research group is currently involved in several microblogging projects including ‘microblogging for logistics’ (think of tweeting RFID chips).

To implement a full ubiquitous microblogging scenario, still lots of work has to be done. Today’s examples from Twitter are individual programmed prototypes. In terms of enterprise-wide ubiquitous microblogging we need much more sophisticated architectural approaches. Currently, we are thinking about how such a ‘microblogging middleware’ could look like.

References:
Weiser, M. (1991). The Computer for the 21st Century. Scientific American, 265(3), 94-104.

I was involved in enterprise microblogging right from its start. Being part of something, it is always hard to have a neutral opinion on it. However, due to my academic view on the topic I would claim to stand somewhat outside the hype centre. So I ask myself these days: what is next?

While others made several predictions yet (i.e. Gartner’s view on the topic) I am especially interested in the level of microblogging adoption. Is ‘microblogging’ really what we see today for example in Twitter? Is that the end of the development? This would mean that the only challenge in enterprise contexts is to adopt it in the right way to create enterprise twitterers.

When I look at microblogging I primary see a huge instrument for information transmission with the recipient choosing the sources. Sources today mostly are humans. There are funny exceptions like the London Tower Bridge (http://twitter.com/towerbridge) or the Tweeting Cat Door (http://twitter.com/GusAndPenny). However, we do not see such implementations in the enterprise. Most enterprise information would fit for microblogging usage but we cannot find it there. Think on new quotes, new orders or new customers (coming from an ERP system), or alerts from the fire control system, or oil level alerts from the company’s cars. There is huge potential in integrating the company’s stream of microinformation using microblogging. Human text messages are only one part of it.

The problem is that in using 100% human postings we started with the most difficult part. Every Twitter user can tell from the problems one have with several 100 followed users. You simply cannot be aware of every posting. However, every posting might be worthy. Therefore intelligent systems should help us to find connections between postings and to filter the most important ones. The bad thing is that this is very hard to achieve in an unstructured 140 character long piece of text. In combining these contents with well-structured streams of machine-readable data we (respectively our computer) could better understand the whole information ecology evolving out of microblogging. I expect the future to bring further developments in this area. Let’s see