By Ralph Grabowski
Issue #745 | August 6, 2012
To see the original article, click here
The problem: the huge amount of money people have to put up front to implement a search solution, such as PLM. But this costs too much and takes too long to implement + the pressure from the vendor to buy more licenses and services..
The solution: an open source enterprise model. Alcove9 does enterprise search for CAD and manufacturing, and has a close relationship with Aras, the open source PLM system.
The technology: Apache’s open source SOLR search engine running an integration framework using Google’s REST APIs. It processes CAD data and images, running inside AutoCAD, Inventor, ARAS Innovator, and so on. Just click the Search button. Or, use it through a Web browser interface. It searches files stored on internal servers, Vault, Sharepoint, and other sources.
Alcove9 CEO Sam Abu-Hamdan told me that he wants Alcove9 to be a one-stop shop. “How can you guarantee this?” I asked. He replied, “We now access text, meta data, and so on in structured and unstructured formats. In the case of proprietary databases, we build connectors using our APIs. We can accesses PMI and other data from CAD models.”
Mr. Abu-Hamdan had such a complete presentation that I could barely think of any questions to ask. One of his slides, for instance, listed his competitors, a topic most other firms I interview are unkeen to discuss. Competitors suffer from these drawbacks:
Inforbix lacks embedded CAD and viewers
Actify is not in quite the same space
Projectwise is very focused on AEC
And similar software from CAD vendors, such as Exelead from Dassault, tends to be focused on the vendors own CAD offerings.
History of Alcove9
The company began in 2004 as Verity, which embedded its search engine MatrixOne. By 2006 120,000 licenses were sold through MatrixOne accounts. In 2007, Autonomy bought Verity at the same time that Dassault bought MatrixOne, and this was the opportunity for the search engine to get into markets beyond MatrixOne.
In 2009, they added unified search, and sold the product through Tata Technologies. In 2011, it was renamed A9 Hub and sold by the Alcove9 company itself. In 2012, it went to open source, and integrated with AutoCAD and Aras Innovator.
Alcove9 plans for the rest of 2012 to enhance the UI of Alcove9, improve semantics, and add video search (in which spoken words are synthesized). In 2013, it plans to release integrations for Synergis Adept, SmarTeam, Sharepoint, and for ERP software, such as Inforbix and SAP.
Customers Using Alcove9
The biggest customer is Honda in automotive, and M7 parts supplier in aerospace. Alcove9 is now entering medical equipment field.
“Where does the name come from?” I wondered. It turns out that the US Library of Congress has eight physical alcoves, plus a ninth one that is virtual.
“What is involved in a typical setup?” I asked. To implement Alcove9, it takes just 5-10 minutes to set up the Web browser, but then depending on the customer’s data, the indexing process can take days or weeks, a process that happens in the background. At a major Japanese automotive company, for instance, it took about a month to get all R&D data indexed and thumbnailed. This is a one-time setup; after this, spiders and triggers do immediate updates of new data.
In the demo that Mr. Abu-Hamdan gave, I was impressed at how easily he could narrow down the search parameters to find specific data. It’s a little bit like using Google Search, in which you can specify “in the last week” and “dwg files.”
You pay for the software through an annual subscription, with a la carte pricing. A gold membership gives you upgrades, helpdesk access, online chat, and email support. Platinum level adds any services you need.