Sunday, May 3, 2015

Innovation – the NTT I3 way

In March 2015, Srini Koushik as the President and CEO of the NTT Innovation Institute in Palo Alto visited us in Munich before his keynote presentation for a customer workshop. I took this opportunity to do an interview on his view on innovation & product lifecycle management.

Srini, please introduce yourself to our readers.
I was both a developer of technology and consumer of technology. My background is more in “applied innovation” than core innovation. My focus has always been on how to use technology to improve things in enterprise settings. I started as a developer at IBM research and global services, became an architect and a distinguished engineer and ended up writing a couple of books, e.g. “Patterns for eBusiness” in 2001. Then I went on to be CIO & CTO at Nationwide Insurance for close to 10 years. The whole focus shifted from developing technologies to applying technology in an enterprise setting.
Srini Koushik, President & CEO of NTT Innovation Institute Inc.

So with the combined role of CIO & CTO, you contributed to the core business of the company?
That’s one of the changes that I was pushing for at Nationwide because I think IT in many places tends to be more of a support function and not integrated into the core of the business. I think for today’s digital businesses, IT has to be part of the core business. That’s a change a lot of CIOs are beginning to recognize. For a company to be successful, IT has to move away from being that order-taker / enabler role to someone who is actually involved with and driving the innovation in the business.

Who is driving innovation for digital businesses – IT or the business?
In the best case, it is a collaborative approach. The big shift I see in successful companies is IT moving away from saying “no” most of the time because of cost, standards, security etc. and helping to drive innovation in the business. On the other hand, the business is realizing they cannot be market leaders without technology.

What is the Mission of the NTT Innovation Institute Inc. (NTT i3, pronounced NTT i-cubed)?
We have 2 objectives: the 1st one is to help increase revenue for our operating companies (NTT DATA, NTT Communications, Dimension Data). We do this by the development of platforms for cloud and security and others. We don’t take these platforms to market directly, but through the operating companies.
The 2nd objective is to make sure that NTT Group is viewed as a global innovator. In order to do that, we have to do the R&D and forward looking research that helps position all of us as thought leaders in the market place.

NTT Innovation Institute Inc.

Tell us about the NTT i3 approach on innovation.
Well, we believe in applied R&D – as discussed earlier - and open innovation. Innovation is not something you can do in one single place, you can’t have a center that does innovation for you. By definition, innovation comes from merging of ideas from different places. Sometimes it comes from our employees, sometimes it comes from our customers or competitors. We make sure that we have a completely open view on where we’re going to get the ideas from.

How do you think about intellectual property protection in this open innovation approach?
All of us who have been in corporate settings have been taught to become convergent thinkers: here is a problem – how do I solve it? You define alternatives and narrow them down to a solution. We believe that innovation is a divergent process. Some of the best innovative solutions come from the combination of ideas, e.g. someone from the financial services industry listening to an automotive expert and modifying the concepts just a little bit for his environment. We believe in going out and listening to as many solutions as possible. Instead of narrowing them down to solutions, we try to build upon them.
Only when you get into the implementation of a specific solution, then you want to protect the IP. One of the reasons why Silicon Valley works so well is the spirit of “no patents in the early ideation phase”.

What else is so special about Silicon Valley as an innovation environment?
A key to success is the availability of experts and their willingness to spend time with each other. There are hundreds of people ready to help you. They may be billionaires who started up other companies and competitors, but everyone will take the time to sit down with you and share ideas. But after you get there, it’s a race. Everyone is trying to figure out how to implement that idea. The ability to execute on ideas is key.
A lot of large companies make mistakes when they setup these labs in Silicon Valley and implement the same management model as in their home countries. They bring their company culture into the valley and expect different results. I think it’s important to recognize that when you’re coming into the valley, you have to have a good mix of people and cultures. People from Germany and Japan have a lot of shared culture in terms of engineering discipline and quality, but this can also get into the way of innovation. Because you deliver quality by eliminating variability, but innovation by definition is challenging this process.

What are the focus topics for NTT i3?
We are focused on applied R&D for cloud, security and network function virtualization. We are not very big, so we focus on research in specific emerging areas such as machine learning and Internet of Things. We actually call it a social network of things because we believe that these devices don’t act in isolation, they actually collaborate with one another and they share information, so it is very much like a social network.

How do you organize your team at NTT i3?
If you truly believe in the divergent thinking approach then you have to build a team made up of people from different backgrounds. So I have a lot of really good scientists, they are experts in security, machine learning etc. But I also have 10-12 people on my team with very little technology background. As an example, I just hired a person with three degrees in nano technologies, so that he can sit down with the rest of my team and they can combine ideas.

How do you transfer your results to the operating companies?
We have different maturity across the globe in terms of using our technology. A key to success is speed to market, so we cannot develop our platforms and then spend one year on training. Instead, we are identifying the areas and people with the right skills to bring our technologies to the market. This is definitely work-in-progress and we appreciate any input from our operating companies to improve this process.

What is your view on the relationship between innovation and product development?
Innovation is a divergent process, product development has to be a convergent process because you cannot work on 20 ideas in parallel and deliver results. The one thing I would like to add to this is that continuous innovation has to be integrated into the product development process. Most of our products are software platforms. This field changes so quickly, that we cannot afford 12-18 months product development cycles, so we have to incorporate continuous innovation into the product development. We have this in common with the automotive industry, they also have to make sure that the product stays current in 3-4 year development cycles. In NTT i3, we use a process called “agile product lifecycle management”. It allows us to constantly look at changes in the market and work that into the overall process. The stage 1 is about ideation and proof of concepts. Once that is done, we create the “minimum viable product” and put it in front of our customers to get feedback. We iterate through this stage 2 frequently. After that, we take the product to general availability. These concepts are not only valid for software development, but can be applied to other product development settings.

Thank you very much, Srini. How can we get more information?
Please visit our website on for access to our publications and views on digital business. I’m also available on Twitter as @skoushik.

Tuesday, May 20, 2014

TCO of PLM – Impressions from the ProSTEP-iViP Symposium 2014

TCO of PLM – Impressions from the ProSTEP-iViP Symposium 2014

On May 13-14, 2014, my favorite PLM conference in Germany took place in the Berlin Congress Center at the Alexanderplatz. 

I’ve been attending this conference for the last 20 years and represent NTT DATA in the ProSTEP-iViP association. This year was another success, with more than 500 participants. Dassault and Brose were the main sponsors, NTT DATA actively contributed with a booth and a presentation.

Total cost of ownership of PLM
In a joint presentation with Airbus, we reported on a project to create a TCO model for PLM. Following the motto “you can’t manage what you can’t measure”, the Airbus PLM TCO model was created in order to steer changes of the PLM architecture by understanding the TCO impact of alternatives. In order to create detailed cost models for PROJECT mode, but more importantly for RUN mode, we reviewed algorithmic estimation models based on COCOMO, e.g. COSYSMO (systems engineering) and COCOTS (COTS commercial-off-the-shelf). These models provide insight into the significant differences between custom-developed PLM solutions vs. COTS. The concept of “technical debt” was presented as a major factor in TCO and obsolescence control. Early investments into PLM software quality as well as preventive and perfective maintenance activities are the main levers to manage technical debt.

ProSTEP-iViP goes global
Although the ProSTEP-iViP asscociation was founded as a German entity, the mission of enabling global collaboration and standardization requires global reach. In 2014, the ramp-up of relationships with Japan is the first priority. A meeting with Toyota on topics such as the JT neutral, lightweight geometry format and the code of PLM openness (CPO) confirmed this priority. As NTT DATA, we are looking into opportunities to support this mission with our Japanese colleagues.

PS: this is a cross-posting from the original post in the NTT DATA blog 

Monday, February 4, 2013

Strategie für PLM-Standardsoftware

Product Lifecycle Management-Plattformen sind – neben ERP und CRM – ein Eckpfeiler moderner Unternehmens-IT. Gleichzeitig hat aktuelle PLM-Standardsoftware einen hohen Reifegrad erreicht, so dass in vielen Unternehmen die Ablösung der ersten Generation von PLM (-Individualentwicklungen) geplant wird. Eine tragfähige Lösung kann nur geschaffen werden, wenn die Spezifika von PLM-Standardsoftware in der PLM-Strategie berücksichtigt werden. So kann mit einer dedizierten PLM COTS Policy (commercial-off-the-shelf) zum Beispiel der große Block der Wartungs- und Betriebskosten nachhaltig gesenkt werden. Oder es können Risiken bzgl. Herstellerabhängigkeit gezielt adressiert werden.

Herausforderungen beim Einsatz von PLM-Standardsoftware

Die Entscheidung für PLM-Standardsoftware ist oft ein Paradigmenwechsel, der als solcher erkannt und gemanagt werden muss. Wenn vorher mit Individualentwicklungen jede Anforderung der kreativen Benutzer erfüllt werden konnte,  ist nun ein ständiger Abgleich zwischen Anforderungen und den Möglichkeiten der Software nötig. Damit diese Kluft nicht zu groß wird, müssen die Ziele und Prioritäten des PLM-Anbieters mit der eigenen PLM-Stategie zusammen passen bzw. passend gemacht werden. Dies ist bei großen Anbietern nicht einfach und erfordert abgestimmtes Vorgehen zwischen IT, Fachbereichen und Einkauf.

Optimierter Einsatz von PLM-Standardsoftware durch eine PLM COTS Policy

Ein Architekturprinzip wie „buy over make“ ist ein guter Anfang. Allerdings muss das Prinzip operationalisiert werden, z.B. durch Trainings für Projektleiter und Architekten oder durch Checklisten für die Betriebsfähigkeit einer Lösung.
Eine gute PLM COTS Policy zeichnet sich unter anderem durch folgende Punkte aus:
-         Strategisches Alignment: durch Deduktion aus der PLM-Strategie entsteht eine PLM COTS Policy, die nachweisbaren Nutzen für das Unternehmen bringt
-         Differenzierung: innovative, wettbewerbs-differenzierende Lösungen können nicht immer mit Standardsoftware geschaffen werden. Eine gute COTS Policy liefert Entscheidungskriterien für Abweichungen von „buy over make“.
-         Integration in EAM: die COTS Policy schafft Wege zur Zielbebauung gemäß Enterprise Architecture Management
-         Integration in Sourcing und VRM: Einkauf und Vendor Risk Management sind erfolgskritische Prozesse beim Einsatz von PLM-Standardsoftware

Entwicklung einer PLM COTS Policy

Die PLM COTS Policy lässt sich in drei Ebenen strukturieren.

Neben der strategischen und der operativen Ebene ist insbesondere die mittlere Ebene „pro PLM Bebauungscluster“ wichtig. Hier werden zum Beispiel Lead-Applikationen für MCAD und ECAD definiert oder die TDM-Strategie für das Simulationsdatenmanagement detailliert.
Als Beispiel für die operative Ebene seien Software-Entwicklungsrichtlinien genannt, die auf die Besonderheiten einer PLM-Plattform wie Teamcenter, ENOVIA oder Windchill eingehen. Durch explizite Richtlinien zu Namenskonventionen, Struktur, Pattern etc. sowie durch entsprechende Check-Tools kann die Software-Qualität sichergestellt werden. Dies ist gerade dann wichtig, wenn das PLM-Entwicklungsteam heterogen zusammengesetzt ist oder die Teammitglieder häufig wechseln.

Haben Sie Interesse an diesem Thema? Sprechen Sie uns an! Oder besuchen Sie uns am Stand auf dem ProSTEP-iViP Symposium am 16. und 17. April 2013 in Hannover. Im Vortrag „Product data is our main asset“ werden wir gemeinsam mit Airbus über Erfahrungen bei der Nutzung von PLM-Standardsoftware in großen Unternehmen berichten.

(Cross posting from NTT DATA EMEA Blog)

Wednesday, May 16, 2012

Terminal Tango: Aerospace and Automotive PLM

Same procedure as every year: on May 9-10, 2012, more than 450 PLM experts met at the
ProSTEP-iViP Symposium in „Terminal Tango”, in order to exchange experiences regarding Product Lifecycle Management (PLM).

As a former terminal of the Hamburg airport, this location was fitting the main sponsor Airbus. This year’s motto was „Managing Globalization - Processes and Systems“. The event provided plenty of information for this: more than 40 presentations in three parallel tracks.

Besides managing out NTT DATA booth, I was able to contribute a presentation together with Airbus’ „Head of PLM Architecture“. We presented our experiences in Enterprise Architecture Management for Airbus PLM.

Airbus PLM architecture in short
PLM architecture at Airbus is complex. After 12 years development time, an aircraft model is produced over 30 years and operated for another 40 years. The product data must be managed over this lifecycle of 70 years – including supplier data in the extended enterprise. The digital 3D model of an A380 (DMU: Digital MockUp) consists of more than one million parts. Different PLM suites for A380 + SA/LR, A400M and A350 are grown historically.

The challenge was to optimize this PLM landscape: current versions of standard software should be used; processes and systems should be harmonized; product data as the main asset should be protected over the lifecycle. On the PLM architecture level, these objectives are supported by using EAM methods and tools. We modeled for example processes, functions and business objects – next to the application architecture – in order to specify Airbus’ requirements for a target PLM landscape. Architecture principles were used to translate the Airbus PLM vision into concrete guidelines and decision criteria.

Aerospace and Automotive
This year was coined by a strong aerospace presence in the traditionally automotive-heavy ProSTEP-iViP association. Besides a dedicated Airbus booth there were 11 aerospace presentations, among them a keynote by the Airbus CIO Guus Dekkers.
There were more commonalities than differences between the two industries. Still, large DMU for an aircraft with more than one million parts has other performance challenges than a digital car mockup with 10.000 parts. On the other hand, aircraft cabin development can profit from customer-centric development methods in automotive.
Question: could you give an example of female-specific requirements in cabin design?
Answer (quote): “how do I know – I’m not female, either” J

The next ProSTEP-iViP-Symposium will be held on April 16-17, 2013 in Wolfsburg, Germany. Volkswagen and Siemens PLM will be the main sponsors.

Monday, February 13, 2012

PLM architecture at Airbus

The global business of Airbus with long product lifecycles poses specific challenges in the PLM domain. In a joint presentation with Airbus at the ProSTEP-iViP Symposium (May 9-10, 2012 in Hamburg, Germany), I will focus on solutions from the PLM architecture perspective, e.g.
-         using Enterprise Architecture Management to harmonize requirements and applications between different national companies and aircraft programs (“backbone approach”)
-         using commercial off-the-shelf software with an out-of-the-box approach to address obsolete components of the current PLM landscape in a sustainable way (“right use of COTS”)
-         modeling of the functional architecture in terms of generic functions and object model to decrease vendor dependence and to protect product data over 30+years (“product data as the main asset”)

The presentation continues to describe how these architecture principles are implemented at Airbus in terms of governance and roadmap planning. Using the example of DMU (digital mockup), the practical application of some architecture principles is discussed. The presentation concludes with lessons learned in shaping the Airbus PLM architecture.

I would love to meet with you at our booth - by then probably under the NTT DATA brand.

Monday, August 8, 2011

PLM Openness and Vendor Risk Management

Have you been following the discussion around PLM Openness? If not, please refer to the article Open words – ProSTEP-iViP Symposium in the BMW Welt, Munich as a starter.

The work on this topic focuses on two action fields:
-         Technical (API granularity and lifecycle, standard support, documentation…)
-         Business & Legal (partnership model, roadmap commitments, cost for toolkits etc.)

I found this second field to be especially interesting, probably because two vendors such as Dassault and Siemens PLM are perceived very differently in terms on PLM openness, although there are a lot of similarities on the technical level.

Enter VRM
IT Vendor Risk Management is an element of enterprise and IT risk management. The purpose is to assess and manage risks related to IT suppliers. Such risk monitoring is mandatory in some industries, e.g. when dealing with personal data (BDSG in Germany), payment card or healthcare information.

A comprehensive VRM approach is based on a rating framework with criteria such as financial performance, staff turnover, customer turnover, product failures, strategy changes or market entry of new competitors. A good question to ask: Would my bank extend credit or invest in the vendor?

Such an approach can be supported with a VRM solution. These applications come as part of a larger enterprise risk management suite or as stand-alone solutions, typically SaaS.

Bringing it together
As technology risk management consultant Joel Lanz points out: „a good contract is the foundation to successfully managing IT vendor risk“. This contract should document all expectations. An extract in plain English should be provided to the team so that everyone knows the rules. A clause defining the “right to audit” supports the monitoring of important contract elements.

What is your experience with VRM for PLM vendors?

Sunday, June 26, 2011

The power of open PLM platforms

While researching on PLM openness, I came across the ebook The Power of Open. It is basically a marketing brochure for the creative commons licensing models, and it contains a lot of case studies where these models were successfully applied, e.g. Wikipedia and TED talks. The term open in this case refers to sharing content through CC-licenses. At first glance, this didn't seem to be useful for my work on PLM openness. After all, we want to protect IP, right?

The case of creative commons was nicely stated on the homepage:
Creative Commons began providing licenses for the open sharing of content only a decade ago. Now more than 400 million CC-licensed works are available on the Internet, from music and photos, to research findings and entire college courses. Creative Commons created the legal and technical infrastructure that allows effective sharing of knowledge, art and data by individuals, organizations and governments. More importantly, millions of creators took advantage of that infrastructure to share work that enriches the global commons for all humanity.

PLM platforms as infrastructure
Open PLM platforms should also provide an infrastructure for effective data sharing. In the best case, this PLM infrastructure can be enriched by 3rd parties, e.g. by adding functionality. This suggests that PLM openness can be applied on data and on functionality - each with an company-internal view and with an external view:

- Complete access to company data
- Open data model
- Support of data format standards, esp. STEP
- API's and documentation for sustainable customization, i.e. supported over multiple versions
- Ability to integrate with other applications
- API's and documentation for 3rd party add-on development including a business model to support this
- Support of service standards, esp. PLM services

Especially the aspect of 3rd party add-ons points back to the Power of Open: the current practices of most PLM vendors seem to focus on developing all functionality themselves. Partners are only used if needed, e.g. when specific know-how for the integration of other applications is needed. The success of App Stores with ten thousands of developers offering add-ons for a basic platform such as the iPad indicates the unused potential of PLM vendors. More openness could help using this potential - what an incentive.