What if… “PLM on the cloud” succeeds?

by Oleg on May 24, 2011 · View Comments

Push aside our fear about security and PLM on the cloud. I was reading Washington Post’s “Video Viewing on Netflix Accounts for Up to 30 Percent of Online Traffic“. The following link leads us picture comparing web traffic coming from various providers. Video is king on the web.

CAD files, visualizations and other heavy weighted stuff. It is not lightweight Twitter 140 chars status updates or Facebook low resolution picture sharing. Cloud service providers from Amazon to Google and Rackspace are building data centers to accommodate the future of cloud applications. It made me think “What if PLM on the cloud succeeds?”, what will be the cost of this solution? CAD and PLM companies are starting to offer solutions to share data on the cloud for better collaboration and data exchange. I wrote about such a type of the solutions before.

The following quote from Wash. Post article is interesting:

Last week, Cable One introduced metered prices for U.S. customers that include 50 to 100 gigabytes per month. According to its Web site, the company (owned by The Washington Post Co.) will charge customers 50 cents for each gigabyte beyond the caps, but it will continue to offer a flat-rate monthly plan also.

Beginning this month, AT&T began limiting data usage to 150 gigabytes for DSL subscribers and 250 gigabytes for its UVerse broadband customers. Users will be charged an extra $10 a month if they exceed the cap. Comcast also has a 250-gigabyte cap for its broadband users.

What is my conclusion? The PLM on the cloud conversations is always about the security and never about the price and cloud usage. I don’t know if my drawings will be stolen faster on cloud. The question what if the transfer of my drawings, models, animations and rest of the stuff on the web will be costly. It might be significantly more costly than today’s software licenses. What is your take?

Best, Oleg

  • Share/Bookmark

Tagged as: CAD, Cloud, On-demand, PLM

  • pradeep

    hi oleg, i do agree but there is a small criteria for this to really happen. well you are comparing things that have totally different infrastructures. PLM products are rigid in terms of infrastructures. the only thing that can really make them available online is to make them customizable and readily available to everyone in addition to data security. even SAP couldn't make it to the top, for all. it really needs an innovation as stated by mr. bernard charles of 3ds.com, that i see what you mean. i presume CAD can be handy in getting the PLM up on the cloud. when users seek and they get easily and also make it to use easily, more and more would flock to it. thats what apple is doing today. they have products that suits not just the generic public, but also a small child who likes to play on its parent's mobile for fun. they have innovated at the level, everyone can use their products in a much simpler way.

  • beyondplm

    pradeep, thanks for your insight! I think, usability and simplicity are two key factors that will play a significant role in the future PLM movement to the cloud. Best, Oleg

  • Hi Oleg - thanks for this article, and for getting the conversation going about cloud PLM. I definitely think cloud is becoming more widely accepted. As consumers become more comfortable using cloud for personal pursuits like email, photo-storage, etc., this will only continue. Also, tools like Salesforce.com have made cloud much more popular/trendy in the business world, even for non-techies, and that will only help cloud PLM. Our VP of engineering actually wrote an article addressing yours, check it out! http://blog.arenasolutions.com...

  • beyondplm

    Alex, thanks for commenting and sharing the link on Arena blog! I agree, salesforce.com made a significant contribution to the awareness of cloud / saas solution. the world was different before sfdc. Best, Oleg

  • pbhalesain

    PLM as SAS will have issues with IP data but PLM on cloud (Which includes in house cloud environments) will not. If we are considering PLM SAS as cloud, then it is mainly a software architectural issue and not necessarily cloud issue. The file data could be stored on In-house servers while the meta data could be stored on public servers. It is architecturally (just a matter of adding a restful file service) possible with "oplm" of which I have recently released a preview version. http://github.com/scientia/spa.... The demo is installed on public cloud at http://sparta.scientia.cloudbe...

  • beyondplm

    Prashant, thanks for your comment and information sharing. yes, I agree - to separate metadata and files can be an interesting approach. However, the need to mirror files can be an important element of this strategy to allow people to access files remotely. I know few PMD/PLM systems that can support it from the technical architectural side. Best, Oleg

  • pbhalesain

    Well Oleg, that is the beauty of open source. you name a problem and we have a solution for it. There are open source dfs which adress such problems. Some are even compatible with amazon S3 api, meanin you could have your application running on public cloud while have your private S3 like distributed file system in-house. Offcourse, it comes with pros and cons.

  • beyondplm

    Prashant, Thanks for commenting! Yes, I'm aware about dfs. Actually, S3 is interesting, but not sure it will be cost effective when everybody will retrieve data from S3. This is more "persistent" rather than operational storage. Do you have any practical experience of running CAD vault on S3. I'm interesting to learn more... Best, Oleg

  • pbhalesain

    Hello Oleg, I wasnt talking about putting the cad data on S3, but more like using HDFS which can be hosted in house. So that the users can upload/download the data from in-house servers while maintaining the metadata on external cloud server. I have never put CAD data on S3 and would never do so.

  • beyondplm

    Prashant, Understood. Btw, what is the problem you see with CAD data on S3? Best, Oleg

  • Sandeep Kongathi

    I don't think PLM Products are matured enough technically to move to the cloud as supposed to Regular J2EE Apps. This might be possible 5 yrs down the line

  • beyondplm

    Sandeep, thanks for your comment! What do you see as a criterion for maturity? What vertical apps (ERP, CRM, SCM, etc.) are ready for the cloud from your standpoint? Best, Oleg

  • I think a lot of this depends. To be honest, I see the value of having the data "in house" and secured behind the firewall. If "just because". I can understand the cloud argument though. The cloud is really just analogous to the electric grid, where at first most large companies had their own power generating capacities, but this is time evolved into the current pay for usage model. This is kind of what the cloud is going after, at least that is how I get it.

    But sometimes having control of stuff just makes sense. I think it depends on the company, and time will tell based on options provided. Part of the issue with PLM would be when you are talking about complicated implementations, hard to set up and hard to maintain. But what if this were easier, if the architecture was straight forward. The OS and application running on a virtual machine instance, snapshotted for backup. The database server on a database server instance, database files backed up by default. And the file vault as separate storage attached to virtual image, setup for replication. The simpler the implementation is, the more encouraging to run such a system "in house" and have control over that system. With virtualization you get better resource usage, at least in theory.

    But then again, what about a hybrid approach? App and database run "in the could", but file vault stored behind the firewall? Database could be encrypted on the cloud, data exchange from cloud to client could be encrypted. And local client uses file location metadata from database to retrieve file locally. Of course, that kind of file vault would need to run a little software to serve up the files (vs. being locally attached storage to OS/application), but that could be pretty lightweight.

    Anyway, I suppose there are many approaches that could be taken.

  • beyondplm

    Eric, Thanks for your insight! In my view, there are 2 things I completely agree - 1/ it depends on a company; 2/ hybrid approach will be the winning one for coming future. In addition, with regards to the electric grid analogy - not everybody can afford generators. It is a matter of cost. So, cloud injection will happen for small customers first. For them cost matters... Just my opinion. YMMV. Best, Oelg

  • Stan Przybylinski

    Hi Oleg,

    I agree that it could be more costly, but I am not sure that data charges will be the only contributing factor. Some are bundling in the savings in IT infrastructure, staffing, etc. into their pricing models to come up with higher usage fees.

    Stan

  • beyondplm

    Stan, thanks for your comment! I think, you are right and moving to cloud has a sense of bundling services. Even so, traffic is a significant part in the cloud business. You can see it also in Amazon EC2 business. Traffic is part of expenses. Best, Oleg

  • I think it's more a case of 'When' and to be honest, the extra bandwidth costs for companies will be nothing to the massive data centre costs that are currently being spend to house internal infrastructures. Like from the transition from the drawing board to the CAD workstation, in years to come users will be saying do you remember when we had to replicate all that data from server to server!

  • beyondplm

    Mike, Thanks for your comment! I agree with your "when" comment. Even so, the massive migration of all data centers to web will cause significant migrations of IT services from home IT to outsource IT. The interesting question is who will be paying an additional cost of services and when it will happen? Best, Oleg

blog comments powered by Disqus

Previous post:

Next post: