As was reported in many German media and in the Hannover Fair’s newspaper on April 14, Angela Merkel was presented at the Fair with the first customised perfume, “Angela’s Dream” produced on a new machine develop in a joint cooperation of Optima with Festo and Siemens.
Optima’s web site gives more background information on this technology, see http://www.optima-packaging-group.de/opg/group/en/news/news_more.php5?id=702
As was explained there, Optima has in the past already produced packing machines for the complete packaging of personalised perfumes. This comprised the ability to select an individual flacon and the corresponding box, which can be labelled individually and is folded directly on the machine.
The new machine demonstrated at Hannover Fair even included a CO2 laser that would engrave the flacon. Visitors at the fair were able to enter their specifications and design their own labels on tablets and “order” with the Optima machine. This gives everyone an individual perfume as far as flacon and packaging is concerned. Welcome do lot size 1 and industry 4.0 as Optima put it.
However, this reminds me more of my younger years when I sometimes personalised Champaign by sticking a personalised inkjet printed label over the original one. That did not change the content of the bottle one bit, and I guess people appreciated that! However, for it to be a true customisation, there is one piece missing in that mass customised perfume production: specifying your own scent. The machine would then really create an individual perfume specifically for you. It is easy to see that it would not be too hard to come up with a machine that does that part of the job.
But creating perfumes is an art form. (That holds for Champaign too!) The challenge would be to create a web interface that allows users to come up with a specification of a scent in such a way, that in the end a perfume can be produced that meets the expectation of the user, given the specification entered. If the user specifies something really awful, then she should at least get a warning. For example the system could describe the expected scent and give tips on how to improve it, for example by warning users if they combine certain components or pointing out that one component would overpower everything else if it’s part was not reduced.
As in all such projects, to build such an interface would need all the help of the experts in the field and a lot of explanation of the software specialists of what can and can’t be done. I remember the hype in the late 1980ies to early 1990ties when everybody tried to build expert systems. These were programmes able to process many rules very fast, and it was then the job of the builders of such expert systems to come up with those rules by interviewing the experts. However, it turned out that’s not how experts work. Rules are for the novices. They are good for a start but once a lot of situations are understood, the practitioners develop an ability to judge a situation intuitively. And that’s when they have become experts.
Experts still know and apply their rules, but more to verify the approach they have to a particular problem. They can’t describe however, how they come to the decision to approach a problem the way thy do. It is my understanding that these problems still exist. Similarly, there is no generic tool that helps you to create a mass customisation user interface. It is really one of the hard parts in most such projects. And quite often, the easier it is to use in the end, the harder it was to come up with.
Any additional ideas or comments? Please feel free to add them below.