Security and permissions are showstoppers to adopt search

June 25, 2014


Search and information discovery is a big deal these days. Inspired by Google and other web search giants, we want information at our fingertips at the right time. I’ve been following topic of search long time. You can jump on few of my previous articles about search – Oslo & Grap – new trajectories in discovery and search; Why engineers need exploratory search? and Pintrest will teach CAD companies to search.

You may think cost and complexity are top problems of search technologies. Crunching lots of data and connecting relevant information requires application of right resources and skills. You will be surprised, but there is one more element that drives low adoption of search in manufacturing companies – security.

Information age articles Enterprise search adoption remains low – survey speaks about survey done among 300 Enterprise IT professionals conducted by Varonis Systems. According to this survey – enterprises are afraid good search solution will allow to people o find information with no permission. Here is the passage which explains that:

The respondents were surveyed at two major security-themed industry events, the RSA Conference in February and Infosecurity Europe in April. When asked to choose the biggest obstacle to enterprise search adoption, 68% cited the risk of employees locating and accessing files they should not have permission to view. Further, even if an enterprise search solution perfectly filters out results based on established permissions, the majority of respondents indicated they are not confident that their organisation’s existing permissions are accurate. Additional obstacles to enterprise search adoption most commonly cited were accuracy of the results (36%), end user adoption (29%) and the ability of solutions to scale enough to index all the data (24%).

It made me think about complexity of manufacturing companies and enterprise organization in general. Established permissions are part of the story. The search results permissions are as good as data that enterprise systems are supplying to search software. GIGO (Grabage in, Garbage out). For many IT organization, management of security and permissions is a big deal. Think about typical manufacturing company. Tomorrow, search system can find all CAD files that were occasionally copy/pasted in different locations and shared between organizations outside of existing PDM/PLM tools. What else, multiple "publishing solutions" created variety of published copies in different formats. Add SharePoint and similar technologies sometimes adopted by divisions against approvals of central IT. Good search solution can be a litmus test to many IT organizations.

What is my conclusion? Manufacturing enterprises are complex. As I described, it driven by strategic, political and cultural lines. Search is disruptive technology that has a possibility to cross these lines and expose many elements of corporate IT problems. So, once more, we learn that only mix of technological and people skills can solve the problem. Strategists and technologist of search vendors should take a note. Just my thoughts…

Best, Oleg

[categories Daily PLM Think Tank]

Dropbox Webhooks and Cloud PDM Pivoting

May 27, 2014


Regardless on what CAD and PDM vendors want, engineers are going to share files on Dropbox and similar file sharing services (Google Drive, One Drive, etc.) Do you remember my PLM cloud concerns and Dropbox reality for engineers post two years ago? 34% of people in engineering departments are using Dropbox to share data. I don’t know what is the number now, but my hunch – it is not going down.

I’ve been reading about interesting functionality added to Dropbox- Webhooks. Navigate to the following Computerworld article to read more – Dropbox plays more nicely with Web apps. The following passage explains what service does:

Web developers can now configure apps to be notified immediately of changes that users make to their Dropbox files, taking some strain off Web servers and potentially giving end users a better experience. The functionality comes via a new "webhooks" API (application programming interface) for Dropbox, which lets developers set up real-time notifications for their Web apps whenever users modify a Dropbox file.

More explanations can be found in Dropbox blog – Announcing Dropbox webhooks:

In general, a webhook is a way for an app developer to specify a URI to receive notifications based on some trigger. In the case of Dropbox, notifications get sent to your webhook URI every time a user of your app makes a file change. The payload of the webhook is a list of user IDs who have changes. Your app can then use the standard delta method to see what changed and respond accordingly.

You can ask me how is that related to cloud PDM? Good question. In my view, this particular piece of Dropbox technology can simplify development of any cloud PDM system. Dropbox developed reference application in the tutorial. Another product referenced in the announcement is Picturelife. Simplify – doesn’t mean cloud PDM will ultimately relies on Dropbox. Many companies don’t want to put their data on Dropbox. Maybe your remember my blog post – How to evaluate PDM before it will ruin your personal productivity? Here is the thing – for most of cloud PDM developers, user experience is the biggest issue. Dropbox is an ideal environment to kick off your cloud PDM development experiments. The majority of companies that not using PDM these days are using Dropbox. For these users cloud PDM on top of Dropbox can be "no brainier". Later on, additional infrastructure can be build and used.

What is my conclusion? To get user traction is priceless. It requires lots of UX pivoting. To find right experience is one of the most critical first steps. Future technologies can be improved and fixed. There are many open web infrastructure these days that can be used to build enterprise products (including PDM). Startup companies can pivot and experiment with user experience with Dropbox based cloud PDM… Actually, established vendors can do the same. Just my thoughts…

Best, Oleg

picture credit

Social PLM and Mobile Dribbling

May 6, 2014

Once “social” was a hot topic for PLM developer and analysts. In my view, the hype went down. Ask PLM people about social applications and be prepared for very neutral response. I asked myself – why so? Social applications can improve the way people communicate and should bring value. Nevertheless, social revolution in PLM is kinda “postponed”. You might be interested – will social apps back, when and how?

My attention was caught by WSJ.D article – Data Point: Social Networking Is Moving on From the Desktop. These numbers made me think about potential benefits between “social” and “mobile” in PLM.


Similar to social apps, the popularity of mobile PLM application is not skyrocketing too. Once excited about the ability to run “everything from iPad”, users got back to their desktops, CAD workstations, BOM Excels and browser applications. Did PLM vendors miss the point of mobile? I asked about that two years ago here. The confusion between “mobile web” and “native app” is probably only part of the problem. When world is going to be even more distributed than today, the efficiency of mobile PLM applications and intuitiveness of how mobile app can present the data becomes absolutely critical. However, mobile app will be used only if it is easier and brings additional value. The best example is taking picture during the presentation and sharing it via Twitters and/or Facebook.

Now I want to get back to social PLM option. I just read about new feature – you can tweet to Amazon to put a specific article or item in your shopping cart. Navigate here to read more. There is nothing very special here. It is all about efficiency. Imagine you found something you want to buy at the time you browse your twitter stream in the morning. To stay in the same environment and put an article to the Amazon shopping cart is all about efficiency. So, here is my guess. Social PLM can reinvent itself via mobile option.

What is my conclusion? The efficient interaction is very important when out of the office and not connected to your well-organized desktop. So, specially designed “social PLM” function can be very demanded on mobile devices. However, the fine tuning of functionality and mobile experience is a key. Efficient user interaction combined together with valuable scenario. This is a key for mobile PLM and social PLM to be successful. Without these two elements, customers will keep walking from social and mobile links to desktop. Just my thoughts…

Best, Oleg

photo source – WSJ-D article

How cloud PLM can reuse on-premise enterprise data?

April 7, 2014


Cloud becomes more and more an obsolete additional word to call every technology we develop I hardly can image anything these days that we develop without "cloud in mind". This is absolutely true about PLM. Nowadays, it is all about how to make cloud technologies to work for you and not against you.

For cloud PLM, the question of secure data usage is one of the most critical topics. Especially, if you think about your existing large enterprise customers. These large companies started PLM adoption many years ago and developed large data assets and custom applications. For them, data is one of the most important elements that can enable use of cloud PLMs.

Networkworld article How Boeing is using the cloud caught my attention this morning. The writeup quotes Boeing chief cloud strategies David Nelson and speaks about very interesting approach Boeing is using to deploy and use on-premise data on public cloud. Here is the passage that outline the approach:

Nelson first described an application the company has developed that tracks all of the flight paths that planes take around the world. Boeing’s sales staff uses it to help sell aircraft showing how a newer, faster one could improve operations. The app incorporates both historical and real-time data, which means there are some heavy workloads. “There’s lots of detail and analysis,” he says. It takes a “boatload” of processing power to collect the data, analyze it, render it and put it into a presentable fashion.

The application started years ago by running on five laptop computers that were synced together. They got so hot running the application that measures needed to be taken to keep them cool, Nelson said. Then Nelson helped migrate the application to the cloud, but doing so took approval from internal security, legal and technology teams.

In order to protect proprietary Boeing data the company uses a process called “shred and scatter.” Using software supported by a New Zealand firm, GreenButton, Boeing takes the data it plans to put in the cloud and breaks it up into the equivalent of what Nelson called puzzle pieces. Those pieces are then encrypted and sent to Microsoft Azure’s cloud. There it is stored and processed in the cloud, but for anything actionable to be gleaned from the data, it has to be reassembled behind Boeing’s firewall.

It made me think about one of the most critical things that will define future development and success of cloud PLM technologies and products – data connectivity and on-premise/cloud data sync. Here is my take on this challenge. It is easy to deploy and start using cloud PLM these days. However, PLM system without customer data is not very helpful. Yes, you can manage processes and new projects. However, let’s state the truth – you need to get access to legacy data to fully operate your PLM software on enterprise level. Manufacturing companies are very sensitive about their data assets. So to develop kind of "shred and scatter" data sync approaches can be an interesting path to unlock cloud PLM for large enterprise customers.

What is my conclusion? I can see cloud data sync as one of the most important cloud PLM challenges these days. To retrieve data from on-premise location in a meaningful way and bring it to the cloud in a secure manner is a show stopper to start broad large enterprise adoption. By solving this problem, cloud PLM vendors will open the gate for large enterprises to leverage public cloud. It is a challenge for top enterprise PLM vendors today and clearly entrance barrier for startup companies and newcomers in PLM world. Just my thoughts…

Best, Oleg

PLM, Mass Customization and Ugly BOM vertical integration

March 19, 2014


A car can be any color as long as it is black. This famous Henry Ford quote speaks about how manufacturing handled customization in the past. That was the era of mass production. The idea of limited customization options combined with high level of standardization and high volumes of batch production allowed to decrease cost and improve productivity. The concept of mass production is applied to different products in process and discrete industries – food, chemicals, fasteners, home appliances and automobiles.

However, mass production is getting less popular these days. What comes next you ask? The next trend in manufacturing is going to be "mass customization". This is the idea of total "custom output". Manufacturing is looking how to create a possibility to produce goods in smaller batches to meet customer specific requirements. Wikipedia article provides a good summary of mass customization concept together with explaining economical value.

The concept of mass customization is attributed to Stan Davis in Future Perfect[2] and was defined by Tseng & Jiao (2001, p. 685) as "producing goods and services to meet individual customer’s needs with near mass production efficiency". Kaplan & Haenlein (2006) concurred, calling it "a strategy that creates value by some form of company-customer interaction at the fabrication and assembly stage of the operations level to create customized products with production cost and monetary price similar to those of mass-produced products". Similarly, McCarthy (2004, p. 348) highlight that mass customization involves balancing operational drivers by defining it as "the capability to manufacture a relatively high volume of product options for a relatively large market (or collection of niche markets) that demands customization, without tradeoffs in cost, delivery and quality".

However, to turn manufacturing from Ford-T production mode to mass-customizable requires lots of changes in the way companies design and build products. My attention caught by McKinsey article – How technology can drive the next wave of mass customization. Read the article and draw your opinion. Author speaks about mass customization trends in manufacturing and how it potentially impact enterprise software and IT. Look on the following picture – the list of "new customizable products" looks very impressive.


New technologies in manufacturing are going to make mass production possible – social and crowdsourcing, customer facing product configurators, 3D scanning, dynamic pricing and many others. Clearly, I can see lots of opportunities in new tech development for software and hardware companies. It also requires structural changes in product development and process organization.

You can ask me how is it related to PLM? I’ve been posting about PLM role in mass customization before. PLM becomes one of the most critical drivers in the way development and manufacturing will be organized. Now, I’d like to be more specific. In my view, it is heavily comes down to the way product information and bill of materials related processes will be managed. The ability to have customer facing configurator, with dynamic pricing, optimizing company manufacturing facilities requires significant vertical integration. Today these processes heavily disconnected and implemented in silos. This is not how things should work in 21st century. To connect custom bill of material with specific engineering option and make product delivery lead time short is an interesting process, communication, collaboration and planning challenge. I found the following passage from McKinsey article connected to that -

True scale in mass customization can only be achieved with an integrated approach where technologies complement one another across a company’s various functions to add customization value for the consumer, bring down transaction costs and lead times, and control the cost of customized production

What is my conclusion? Mass customization ends up with ugly bill of materials (BOM) integration challenge. By enabling BOM vertical integration, future PLM systems will make mass customization processes possible, shorten time from the moment customer hits company e-commerce web site and until the moment, product will be shipped. Just my thoughts…

Best, Oleg

[categories Daily PLM Think Tank]

Future CAD Platforms and Google Chrome Native Client

September 20, 2013

Our life is getting more and more web-like. Think about applications and tools we use in our everyday life 10 years ago and now – you can see how many of them moved from your Windows desktops to web browsers and mobile devices. However, if you are engineer using CAD application and/or simulation tool, most probably, you are still anchored to your desktop machines. The same you can probably say about photo and video editing applications. The common thing between CAD and photo / video editing is related to the need to use extensive computation and/or graphic resources.

Speaking about photo editing applications, Google is clearly making a leapfrog activity in this space. Google+ photo editing application is getting better everyday. Many times in my personal life of photo hobbyist I ended up with editing photos using Google+ without reaching to my usual Photoshop tools.

I’ve been reading TechCrunch article earlier this week – Google’s Bet On Native Client Brings Chrome And Google+ Photos Closer Together. This article confirms my guess about Google technologies behind new Google+ photo editing tools as well as made me think about some potential opportunities in CAD / PLM space. Here is an interesting passage from the article.

As you’ve probably heard a thousand times now, it’s virtually impossible to build great photo apps that can rival the likes of Photoshop in HTML5. That’s where Native Client comes it. This technology allows developers to execute native code in a sandbox in the browser. It can execute C and C++ code at native speeds and with the ability to, for example, render 2D and 3D graphics, run on multiple threads and access your computer’s memory directly. All of that gives it a massive speed bump over more traditional HTML5 apps.

If you want to learn more about Google Native Client, you probably can start here. Google Developers website provides a good set of information well organized with use cases, videos, documents and references. Navigate here to read more.

It is interesting to see common use cases presented on Google Developers website. Some of them are very relevant to CAD / PLM domain – enterprise applications and legacy desktop applications. Another interesting use case is related to existing software components. You may think about Geometric modelers as one example of existing components that can run inside of Google Native client. Look on how Google phrase this use case on the development website:

Existing software components: With its native language support (currently C and C++), Native Client enables you to reuse current software modules in a web app—you don’t need to spend time reinventing and debugging code that’s already proven to work well.

Compiling existing native code for your app helps protect the investment you’ve made in research and development. In addition to the protection offered by Native Client compile-time restrictions, users benefit from the security offered by its runtime validator. The validator decodes modules and limits the instructions that can run in the browser, and the sandboxed environment proxies system calls.

Let me speculate a bit here – recent announcement of Siemens PLM about licensing of Parasolid components to Belmont Technologies developing cloud CAD can provide a potential use case. So, maybe future cloud CAD of Jon Hirschtick with use Google Native Client… who knows?

The following video provide you short summary of how Google Native App works.

What is my conclusion? Web is a future platform for everything. Engineering and manufacturing applications are not exclusion from this rule. However, it will not happen overnight. Companies made significant investment in existing technologies and products. How to move from today’s mostly desktop CAD into future cloud design platforms? This is a good question to ask CAD technologists, industry pundits and internet developers. Google Chrome Native Client provide an interesting technological set to consider. Today Google Chrome Native apps directory contains only games. But who knows what will be tomorrow? Just my thoughts…

Best, Oleg

Who should be the first PLM user in a company?

May 9, 2013

Enterprise software implementations are usually not a simple task. Compared to selection of your next mobile device and RSS reader, it is an organizational effort. Enterprise software gets really complicated when it comes to the point implementation requires involvement of people. Product lifecycle management (PLM) is one of these systems. Implementation of PLM is deeply connected to product development and manufacturing processes. Success or failure of PLM implementation is directly impacted by how people are involved in PLM system adoption and use.

Companies are taking different approaches in implementing PLM. However, fundamentally, I can see two different ways in implementation. First is holistic approach usually called "business transformation". It implies significant process changes as a result of PLM system implementation. Companies are analyzing their existing processes, optimize and restructuring the way they do business. Second approach is focusing on a specific process or problem solving. It is usually come as an improvement of a specific activity and/or process.

There are lots of debates about PLM implementations these days. The value of PLM system implementations becomes clear to organizations on different levels. At the same time, it is obviously not easy to people to understand how to start using a PLM system that will have such a significant impact of everything they do.

I was reading an Minerva blog post – Should we pull PLM deployment? A new lean deployment strategy by Yoann Maingon. In this article Yoann shares his view on different approaches to implement PLM. The idea of lean and "pulling data" resonated. Here is an interesting passage:

The lean concept is highly based on a pull flow. Most of the arguments I’ve had were about the fact that the main data is created in Engineering so we should start deployment in engineering. Well, what if you should provide a system to the first person who enter the system. The one who will pull the flow, the customer? the marketing? assistance & support?

It made me think about how to maximize the value of PLM implementation withing short period of time. Here is the idea. Every company is manufacturing products for customers in some ways. The biggest process loop in every manufacturing company starts from requirements and ends with "release" of product to customer. To control the loop between requirements and results can be an interesting problem to handle first.

The idea of "pull" will be related to pulling of product requirements and documents representing released products and combined them together in a single system. In my view, it can provide an interesting insight on company operation. It is also very useful information source that every company can "re-use" for different purposes – new projects, customer support, etc.

What is the conclusion? It all starts from ROI. How to make it faster… This is a challenge most of PLM implementations are facing these days. For most of the implementations the process of getting to results can be slow. To provide system that can capture requirements to release control can be an interesting option. Lots of valuable information is hidden in this relationships of requirements-result. It also can drive management attention and focus in a company. Just my thoughts…

Best, Oleg

The future of PLM Glassware?

April 19, 2013

Technological predictions are tough and nobody wants to make them. Back in 2010, I came with the following post - Who Can Generate 3D/PLM Content For Apple iPad? Back that time, the value of iPad was questioned by many people. Speaking about manufacturing companies, people were very skeptical by the ability of iPad to bring meaningful functionality.

Fast forward in 2013. I’m sure you’ve heard that Google Glasses are coming. If you don’t know what is that about, navigate your browser here. The initial leaks of Google Glass experience are starting to leak the blogosphere. Navigate your browser to the following article - Google Glass is finally here: Tech specs released, first units shipped. The article put some snippets of Glass specification such as display resolution (similar to 25″ monitor), 16 GB flash storage, 5M camera with option to create 720p videos and sync to 12GB of Google cloud storage.

Google is paying a lot of attention to developers community. Navigate to the following article - Google publishes Glass’ Mirror API preview for developers. It contains a link to Google Glass Mirror API you might find useful.

Here is another article that caught my attention - 10 things about Google Glass: Could this be Google’s iPad? Many of Glasses usages are clearly individually oriented. At the same time, author is raising some initial questions and thoughts about business usages of Glasses and how Glasses can be connected to corporate accounts. Here is an interesting passage:

Consumer Google accounts can be connected to Google Glass. No corporate connections yet. The real interesting connection for enterprises would be service-oriented businesses and Google Glass. For now, Google Glass is all about individual accounts. Google Apps access will certainly follow at some point. The business implications for Google Glass will appear later. Google Glass could become a productivity tool. Presentations, location data, sales information and real-time information on the go could be handy. You could also picture a person on an oil rig giving a real-time, real-world view of a product to a manager in Dubai.

Few years ago, Microsoft and BMW released a video – Manufacturing Future Vision. Watch it below. You will find funny, but many of concepts related to tablet computing world are actually reality now.

However, I want you to pay attention to few examples below very similar to what we can see in a futuristic videos of Google Glass interface.

What is my conclusion? The analogy of Glasses with iPad is very strong. Only few years ago, iPad was introduced a complete new experience. Now, we can see tablet computing experience in our everyday life everywhere. Business usage of tablet computer is skyrocketing. I can see Glass experience can change some of businesses as well. We are going to see many Glasware use-cases that will change company processes. Just my thoughts…

Best, Oleg

PLM and Engineering Task (Process) Management

December 13, 2012

PLM is all about process management. This statement comes to the play when people explain the value of PLM in the organization. Usually, when you think about process management, your mind is switching to some kind of "workflow thinking" mode, which assumes you need to follow the process from state to state by accomplishing tasks and activities. In every PLM implementation, this is a moment of time, people ask – how do we manage engineering processes? What toolset we need to have to make it happen?

I can see, engineering people, are bad organized. In many situations to run processes among engineers is similar to herding cats. To manage process in an engineering organization is a challenge. This is a place where PLM vendors usually fails to provide a reliable and simple solution. Engineers are asking for additional flexibility and vendors have a tendencies to provide a complicated solutions. Many PLM tools are providing sort of Workflow designer to create a process model. Later on, you can discover that engineers tend to abandon these processes. Main reason – these processes are not reflecting the reality. I wanted to come with some ideas how to fix that. I came up with the three definitions – tasks, engagement and information context. Take a look on the picture below.

The overall engineering process is described as list of tasks (above). This is the simplest way to present what needs to be done. It easy to digest and follow up. At the same time, the activity around this task list is not linear. In order to accomplish the task, an engineer needs to engage with additional people. This a typical situation when a person who leads the process needs to communicate with other people and comes with the result. Often, it is ad-hoc communication that cannot be formalized resides in people’s mind. Another situation happens when an engineer needs to bring an additional set of information to accomplish the task or make a decision. To combine these activities together is not a simple thing. Workflow is a wrong tool to solve this problem. To support a simplified task management tools with the ability to manage external engagement and connect to information context can be a potential solution to the problem.

What is my conclusion? The simplification is a key word to summarize my thoughts. In many situations, engineers will prefer a simple task list to get things done. However, tools need to provide a collaborative capabilities to connect the engineer’s activity to other people and additional sources of information. Just my thoughts. I’m interesting to learn how you manage engineering tasks in your organizations.

Best, Oleg

[categories Daily PLM Think Tank]

PLM Innovation: Who will provide PLM to Boeing in 2015?

October 25, 2012

I’m in Atlanta these days attending PLM Innovation Americas conference. Today is the first day, and I’m still expecting many things to come. Nevertheless, here is the thing that made me to blog. I’ve been attending first keynote made by Kevin Fowler or Boeing Commercial Airplanes . The name of the presentation – The sky’s the limit? Reaching for the value of PLM at Boeing Commercial Airplanes. The presentation provided a very interesting retrospective on how PLM developed at BCA (Boeing Commercial Airplanes) for the last 20 years. In my view, Boeing is one of the companies that stands at the root of PLM development back 25 years. Boeing fundamentally influenced the way PLM tools were developed by vendors like Dassault Systems and Metaphase (now Siemens PLM).

Growth of PLM system was driven by the need to support development and manufacturing processes. The ultimate request was to provide additional features to product development processes. Boeing moved from spatial integration (777) to build integration (737-next-gen) and functional integration (787). It means more and more features. The following slides show a very interesting trend related to Boeing PLM Value.

As you can see, functionality is growing. However, total cost of ownership is growing too, which is not very good sign. The most disappointing factor is related to usability of PLM systems. The usability of PLM tools is going down, which confirms the fact PLM is struggling from complexity and user experience. The level of information complexity at Boeing is impressing – 1B (Billion) parts to manage during the release of aircraft, Boeing has 12000 aircrafts in service and operation, Boeing employees are sending 63M (Million) emails during the month. Boeing engineers need to use multiple tools during their working day.

However, the most important message I captured from Boeing presentation was related the future evolution of PLM system. According to the Boeing, all PLM systems Boeing is currently running are approaching their end of life. Take a look on the following slide presenting four generations of Boeing PLM systems. Actually, it represents only three generations. Boeing put a question mark on the generation of PLM tools that need to come in 2015.

What is my conclusion? It is very interesting to see how large companies are turning towards usability of PLM tools. In the past, usability wasn’t a big deal for enterprise companies. Nowadays, end users cannot tolerate bad user experience. According to Boeing, their users are opening about 19 user interfaces to get data. Future PLM systems need to come with a different level of usability and user experience. The expectation level will be set by iPhone, Google, Facebook and other consumer tools. PLM vendors need to make a note. Important. Just my thoughts…

Best, Oleg


Get every new post delivered to your Inbox.

Join 239 other followers