Will PLM Move Beyond Search?

plm-beyond-searchWhat is the role of Search in PLM? This is question I want to discuss today. This year is busy for search community, Microsoft Bing and now Microsoft Yahoo agreement. I had chance to hear so many opposite opinions about search, so I decided to put this question in PLM Think Tank. On the one, very extreme side, I read Stephen Arnold is saying – search is a commodity or give away for free. On the other side of the street vendors like Autonomy is creating OEM alliances with Siemens, which include Siemens PLM business. I’m sure will not be able to put all opinions about search in this blog post, therefore, will be focusing on areas of search that relevant to what PLM is trying to achieve in organization.

I think for many years, Search was considered as simple function in CAD, PDM and lately Product Lifecycle Management interface. This is not true anymore! I’d expect search will start to play much more significant role in everything that happens in enterprise space. There are few reasons for that. First, search was very successful during that past few years in consumer’s applications. Second, enterprise landscape becomes extremely complex. Third , enterprise vendors will look for differentiation and search will be one of them.

These days, search vendors are trying to find special way for their solutions to enterprises. Some of them, I’d like to mention here:

Social Search. Since everything social becomes very popular, search is also here. There are many vendors focusing on searches in social networks. The key focus of these products is to insure real-time search capabilities in changed social networks. I don’t think real-time is extremely important for product lifecycle, but to be able to resolve complexity of social relationship in enterprise can be very interesting and beneficial.

Enterprise Information Search. In traditional approach, enterprise search vendors were focused on indexing of documents in enterprise content management system, enterprise vaults in beyond. Today’s trends show more interest from these vendors to go beyond document indexing and to show more functions to aggregate information and present it to the users.

Visual Search. You better will see it once… Yes, to be able to see results visually becomes important. I had chance to speak about 3D searches, and visual searches on my blog before. There are multiple aspects of visual effects and capabilities. From being able to see 3D models, to visualize graphically results and dependencies.

So, what is my conclusion so far? Success of search in a consumer market will continue to put demand from customer’s side and drive vendors to put more energy to search space. The solution is not there yet.

Best, Oleg.

Share

Share This Post

  • Mohana Krishna Ram

    2D/3D Geometry Based search integrated with Windchill PLM is now available in the market.

    http://www.sconce.sg/pdfs/Bingo.pdf

    An innovative component search solution for the engineering world using 2D/3D shapes as input

  • Mohana, Thank you for link. I will take a look. I understood you focus on shape-based search. I had chance to talk about it before http://plmtwine.com/2009/04/28/what-is-the-future-of-search-for-plm/;

    There are few other solutions that I had chance to see in this field –
    http://www.ingeneastore.com/PartBrowser.html
    http://www.iseekcorp.com

    My early conclusion about 3D search was about fact, that nobody is interested in “just 3D search”. Normally, this is something that can be combined with other search capabilities. I had good conversation about that with Andy on this post – http://plmtwine.com/2009/05/22/top-five-disappointing-plm-technologies/

    Again, Thanks for your comments!
    Best, Oleg.

  • ‘Search’ is quite a loose term. I think there are many information retrieval tasks on product data which aren’t well supported in current PLM software. As someone said to me recently, we’ve got all this data in our PLM system but it’s like it’s not plugged in properly.

    I think you’ll only have limited success in supporting these tasks by bolting on generic enterprise search software – essentially focused on building inverted indices from text-based documents. Product data is structured differently and results need to be presented differently.

  • Andy, thanks for commenting. I’d imaging search solution is complicated. Enterprise search have very limited capabilities when it comes to enterprise systems. Could you explain what do you mean by “it’s not plugged in properly”? Thanks, Oleg

  • Oleg, the quote was from a guy who was running a PLM installation for several hundred users. His basic point was that his PLM data was a hugely valuable asset. Their system was great at data management. However, there were many times when there was information they were looking for, which they knew was in the PLM, but which was very hard and time consuming to get at. Usually, they just wouldn’t bother.

    He gave a number of specific examples. Some were ‘search-y’ i.e. find a specific piece of information given incomplete knowledge of what you were looking for. Others were about aggregating data: ‘how many times have we…’. Others were more like data-mining – finding patterns in the data.

  • Andy, thanks! This is make sense. I think Lock-y behavior of PLM systems is back side of their claims to be able to manage product data in the way nobody else can. This is right for certain degree, but very dynamic, in my view.Regards, Oleg.