Welcome to the world of Infosys Engineering! It is a half a billion plus organization that takes pride in shaping our engineering aspirations and dreams and bringing them to fruition. We provide engineering services and solutions across the lifecycle of our clients’ offerings, ranging from product ideation to realization and sustenance, that caters to a cross-section of industries - aerospace, automotive, medical devices, retail, telecommunications, hi tech, financial services, energy and utilities just to name a few major ones.

« January 2013 | Main | April 2014 »

April 30, 2013

Automotive Telematics - What Lies Ahead

Telematics typically refers to the integrated use of Telecommunications and Informatics, also known as ICT (Information and Communication Technology). Though Telematics has found applications in a number of domains, automotive telematics still remains one of the most prominent and promising areasof its application. As per Machina Research, 'From less than 90 million connections globally in 2010 the automotive M2M market will grow to almost 1.4 billion connections by the end of 2020'.

Automotive telematics consists of various technologies working together in a typical M2M scenario. The technological advances over the past decade related to communication (3G/4G), GPS, Cloud, Mobility, Big data, Social networks etc have really opened the doors for innovation in this area. The kind of applications and products which are possible using the abundance of information which can be tapped from vehicles is only limited by our imagination. A vehicle typically generates diagnostic data related to the engine, fuel, brakes, gears, location etc. This data is relayed to the backend server which processes it and mines the relevant data which can be used in various applications.

Some of the most typical applications of Automotive telematics consist of vehicle tracking, geo-fencing, fleet management, route management, cargo management, vehicle diagnostics etc. These applications are available in the market since a long time and are provided by multiple vendors, OEMs or third party application providers. While such applications will continue to drive the automotive telematics market, the road ahead clearly shows the move towards data analytics. This is accelerated by the parallel emergence and adoption of big data technologies and predictive analysis techniques.

The vehicle insurance sector is taking a keen interest in the information that can be tapped from vehicles and the people driving them. There are companies which process the huge amount of data coming in from vehicles and do predictive analysis regarding the vehicles and the driving habits of the person driving the vehicle. This in turn helps the insurance companies to price their insurance products accordingly or even offer usage based insurance options like Pay as you drive (PAYD) and Pay how you drive (PHYD) etc. Insurance premium can be based on the driving habits, route taken, weather conditions, traffic situation, speed of the vehicle etc.

Another growth area is seen in the social marketing area. The data coming from the vehicle can be mined to understand the characterestics of the driver, his likes, dislikes etc and based on the location of the vehicle and other parameters, targetted advertisements can be sent to the drivers. This has huge potential in the advertising and marketing business. Imagine that you are driving towards the city and the vehicle reminds you that it is your sons birthday next week and you should buy a particular toy from a particular store since they are running a clearance sale, or if you are driving through the city and you get a message saying that your favourite movie is playing nearby and you can book the tickets by just responding to the message. We will get more used to machines giving us suggestions based on our profile and behavior.

Apart from these analytics based use cases, there are other areas in which telematics will bring in innovations. One of the key areas which is being targetted by multiple OEMs is security. There is a lot of research going on to create an emergency assistance network on the back of regulatory mandates like the eCall in Europe. Vehicle Infotainment is another promising area which will allow content providers to stream the content to vehicles over high speed communication networks.

Machina Research expects the automotive M2M sector to generate EUR157 billion revenue in 2020, up from EUR10 billion in 2010. The biggest revenue generating segments will be security and tracking, emergency assistance, navigation and insurance. Automotive telematics offers great opportunities for vehicle OEMs, service providers, vendors and third party integrators. There is potential for growth in the B2B, B2C areas based on the onboard telematics or backend vehicle platforms. We are at just at the tip of the ice berg right now, the possibilities are infinite.

April 5, 2013

Enable your application for IPv6

As per data collected by Internet society, IPv6 adoption is gaining momentum, across the globe. Though IPv4 is not going away anytime soon, it is clear that IPv6 adoption is on the rise. This makes a good case for software applications, which follow client-server architecture and use TCP/IP based communication, to enable themselves to communicate over IPv6 protocol. Abundant text is already available about why IPv6 is not just about overcoming IP address space crunch and how it is a more efficient protocol than IPv6. In this blog, I explore the challenges associated with changes needed at the application level, to enable support for IPv6. Legacy applications may not have been written anticipating the need to change the communication protocol and hence the complexities of identifying the impact areas. Most modern operating systems are dual-stack i.e. they support both IPv4 and IPv6. The application layer should make use of this capability and add support for IPv6 to existing codebase instead of replacing IPv4 completely with IPv6. This approach will ensure everything does not switch over to IPv6 overnight. Below mentioned is a stepwise approach to add IPv6 support at application level:
Assessment: The first step is to identify impact on socket communication between client and server, data exchange between various components and UI.
A) TCP/IP communication:
To facilitate determination of impact on socket-level communication, Microsoft provides a command line based tool, named Checkv4.exe. This utility inspects the code and points out code constructs which need to be modified to support IPv6 and provides suggestions on exact code change too. Though this utility works on one file at a time, it can be easily invoked in a program that iterates over all the files in a project or folder hierarchy and invokes Checkv4.exe on each file. For our purpose, we had created such a program and also captured the output of Checkv4.exe for each source/header file into a text file. This is useful to size the amount of change needed and to arrive at necessary effort estimation for the changes. Checkv4.exe helped us find out relevant APIs and data structures in C, C++ and VB code. Checkv4 is suitable for windows-based codebase. For non-Windows source code, a run-time utility IPv6 CARE can be used to locate impacted areas.
As mentioned earlier, most modern operating systems allow IPv4 and IPv6 to co-exist, recommendations provided by Checkv4.exe are on the lines of replacing IPv4-only APIs and data structures with IP-agnostic counterparts. The IP-agnostic APIs and data structures abstract the differentiation between IPv4 and IPv6 address from invoking programs. A detailed guide to IPv4-only data structures and APIs and their IP-agnostic counterparts is available over here.
If there is a single component which handles all socket level communication, scope for these modifications is limited to this. If the communication code is scattered across different components then it is a good idea to create a new component whose sole purpose is to take care of communication (This is per the widely known principle of 'Separation of Concerns')

B) Data Exchange:
The application suite that I worked on involved multiple components written in different programming languages. Some of them were related to UI while others contained business rules. For intercommunication, these components exchanged user-defined structures and their collections. As the components had been implemented for IPv4 addressing scheme only, the data structure elements were of a size that specifically catered to IPv4 addresses, and hence were incapable of holding IPv6 addresses of longer length. The easiest way to locate such changes is to trace the data flow from UI till the data is sent to server and identify the containers that carry data. Enterprise Architect by Sparx Systems can help generate control flow and UML diagrams from application execution. The approach taken was to add new members to hold IPv6 address and retain the IPv4 data member too. Intent behind retaining data member for IPv4 was to allow socket communication in scenarios when either of the client or server modules was hosted on an old operating system which supported only IPv4 stack. User provided values on GUI and data received from server were validated (there are in-built routines in most programming languages to verify addressing scheme) to determine whether which addressing scheme was in use. This helped in avoiding usage of conditional compilation symbols which reduce readability of code.
C) User Interface for IPv6:
Though IPv6 address is hard to memorize and key-in, applications still need to support IPv6 addressing in GUI controls. To allow ease of search, reading and general usage, rules mentioned in RFCs need to be adhered to. RFC 5952 deals with UI representation of IPv6 address in a very detailed manner. Compliance to the RFC standards ensures application behavior is universally consistent and ensures ease of access for users. Winsock is a UI control built into Visual Basic UI designer which has been traditionally used for displaying and validating IPv4 address. In VC++ based GUI applications, masked edit control has been put to good use for the same purpose. Thus, these two could be used as starting point in GUI layer, to determine impacted areas.
After the assessment of modifications needed is complete, overall approach for change can be designed and effort estimates drawn.

Adopting Big Data - Challenges and Success Factors

Big Data comprises of 3-V factors; namely Volume, Velocity and Variety.  However, considering well understood benefits reaped on adopting Big Data in enterprises, one could be tempted to club another -V (Value) to the existing troika. Mckinsey Global Institute Big data study says that 'The total amount of data created and replicated in 2009 was 800 exabytes -- enough to fill a stack of DVDs reaching to the moon and back' (source: Mckinsey global institute. Big Data: The next frontier for innovation, competition, and productivity. May 2011). Though Big Data adoption is well established in industry segments like Retail, Financial and Insurance, and Manufacturing etc, there is still a need to continuously innovate and implement factors that guarantee success and enable rising returns. The primary challenge in Big Data implementation is the need to handle high voluminous data (that exist in multiple storage sources and multiple data formats ) at the speed it is received and processing to generate intelligent business insights. This leads to an even more complex problem to solve - that of management aspects, where the enterprise structure and processes will need to change in response to findings from Big Data analysis to enable the enterprise to evolve and reap business benefits.


Some of the technical challenges to Big Data adoption include the following :

  • Choosing the right technology fit for processing and analyzing Big Data and providing in time analytics and intelligence.
  • Coming up with extensible and scalable architecture to support current and future needs economically
  • Focusing on operational aspects
  • Focusing on ease of use

Similarly challenges from the management aspects include: 

  • Consolidation of data at organizational level rather than limiting it to internal business unit level
  • Defining and categorizing short term goals and long term goals to aid in business decisions
  • Identifying and involving right skilled data scientists who can look deeper into data and provide business intelligence and provide valuable recommendations. 

The above challenges are generic and are applicable in almost all industry segments though not at the same level. In spite of these challenges, leveraging Big Data for better business insights become a key basis for any firms to cope up with competition and growth. While those challenges have to be addressed by adopting right strategies and involving right stake holders and expertise etc, it is important to understand critical success factors that realize the real potential of Big Data. These success factors vary depending upon industry segment, business objectives ,enterprise structure etc and hence have to be addressed accordingly.

In the manufacturing context, data is available in various formats and across various business units. Consider data in terms of a Product Life Cycle - there is data related to product engineering, product design, product quality, product manufacturing, partners and vendors specific data etc. This data could be stored in various formats and at various destinations. One needs to extract value from a large volume of such data by getting a holistic view of this data available from these various sources, analyzing as per the business objectives and obtain fine insights that will improve product quality, optimize product design/manufacturing cost, understand various trends to improve product value etc. Looking ahead, all the data related to product engineering and manufacturing, when clubbed with sales and support data, can give important insights that would increase sales and help beat competition.  So, it is important to get visibility of all available data and understand various trends to achieve additional business insights.

Some factors that enable extracting real value out of large volumes of data are as follows :

  • Understanding data and mining it efficiently: Identify and consolidate scattered data to enable processing that enables Business intelligence.  Compared to hitherto existing data warehouse and BI solutions and approaches that enterprises adopted for business intelligence, Big Data solutions can contribute much better value to improving an enterprise's business by analyzing data in both structured and unstructured forms to extract meaning. But, it is critical to know how to use this data, what to analyze to come up with real values and right decision support system.
  • Choosing right technologies and architecture: Have right technologies for handling multiple types of data, ranging from structured to unstructured to semi structured and to support the scale and speed at which data is received.  Though technologies like Hadoop and NoSQL data bases like Mongo DB, CassandraDB, CouchDB are synonymous at present with Big Data analytics, there are number of other technologies from commercial vendors like SAP, IBM , Oracle, and Microsoft etc and many other open source technologies such as R, Cascading, Scribe, Elastic Search etc as well.  As we know, one size doesn't fit all, and hence technology evaluation for specific needs has to be done before choosing the right tools and technologies. Apart from this, coming up with right solution architecture becomes critical as the data volume, data types and rate at which data is received is very high. It needs thorough knowledge on capabilities and limitations of NoSQL data bases, distributed computing technologies and cloud computing. Solution should be scalable and extensible to support future demands. Factors on maintainability and support should be considered as well.
  • Data Scientist Expertise:  As compared to traditional expertise on technical and project business side, for reaping the real benefits of Big Data, enterprises have to have right skill sets. Data scientists can help understand the breadth and depth of the data from the business/domain aspects and examine the same to derive various trends and come up with innovative perspectives, which would support newer business opportunities and growth potentials. It is critical to identify and involve the right combination of skills on domain, data, technical and management side to make the Big data initiative successful.
  • Flexibility for Cultural change:  There is a need for cultural change in the organization for making Big Data initiatives successful.  Leaders have to educate business group on challenges and put a strong process and governance for enabling better collaborations. Specific issues like data security and data ownership have to be addressed convincingly. CEOs have to focus on value of Big Data and identify right IT executives who can execute the strategies well. Depending upon the available opportunities, a flexible common platform which enables easier access for infrastructure setup, data analysis, work load monitoring and getting business insights should be envisioned and implemented at enterprise level. Each business units can leverage such platform to analyze specific data and get required advantages more economically and efficiently at business unit level as well.

Above success factors are applicable to any industry, but it is important to understand all these with relevance to corresponding industries and objectives and put a robust process as per enterprise strategies.


April 1, 2013

Off-board Telematics Services in India

India's rapid economic growth over the last decade, emergence as a leading producer and exporter of cars and a large population of mobile phone subscribers all point to a potential lucrative market for telematics (high reward, high risk). It is early times yet as compared to telematics service adoption in the West, but players have already started making the right moves to gauge adoption potential. Our recent interactions with vehicle OEMs and suppliers seem to indicate the beginnings of a rat race to building telematics platforms and innovations around services on those platforms. Make no mistake - the Indian market has already been exposed to basic telematics services like GPS-enabled navigation and vehicle tracking systems over the years, but adoption has been niche and isolated.

Until now, we have been slow .....
Unlike in the west, the Indian market has been slow to adopting telematics services. India still lacks a developed and disciplined road infrastructure. Besides, newer and newer neighborhood short cuts are invented everyday - too many for maps to remain updated. Drivers do not need navigational tools when they could always tap into co-travelers or passers-by for directions that are more accurate. Also, rural regions are not part of the digital mainstream (in terms of mapping locations) and hence, navigational aid will only be restricted to cities. Besides, as in most emerging markets, cost has been a major factor and buyers are sensitive to paying more for services they presume are not that important - having not been exposed to this before. In developed markets like Europe where driver salaries, fuel costs and other operating expenditures are high, savings from telematics services are much more appreciated than in typical emerging markets where these costs are comparatively low. However, with fuel costs spiraling over this year, consumers in India are starting to be fuel-conscious too, and hence services that enable optimization of route, detection of fuel pilferage, detection of inefficient driving behavior etc. will be sought after.
Telematics adoption is highly fragmented in the passenger car and the trucking segment (single truck owners might find the cost of installing telematics devices and subscribing to services too costly as their margins are low - as compared to truck fleet owners) while in the commercial vehicle segment , fleet operators, logistic companies and taxi companies are realizing the benefits of adopting telematics value added services like monitoring and routing services, package tracking, accident management, driver management, fuel management, fleet management etc.

.... but there is potential
India represents the world's second largest mobile phone market and the world's third largest mass of Internet users. Rates for data and talk time being the cheapest in the world, this could be reason good enough to hasten adoption of SIM based on-board units for communication with the back end over 2G/3G and 4G networks.

...and things are changing
With widespread smartphone usage, consumers are today conscious of the advantages of application services on smartphones today. Consumers are hence expecting similar experiences on their vehicle units whereby they could hitch their phones to the car unit and access phone services on the car unit or something similar (such facilities are good as part of vehicle-infotainment solutions, but for true vehicle telematics this may not be much relevant beyond navigation services.) Vehicle telematics will require an on-board unit on the Vehicle CAN bus that can capture a lot of relevant information and convey data to the offboard server. Per an article here, last year's sales figures show that only 7,000 of 2.6 million new cars sold in India came with an embedded navigation system. In India too, recent spiraling fuel process and rising concerns of environmental sensitivity are forcing people to be fuel-conscious and telematics services that enable regulate fuel consumption without being in-car, may well be appreciated. People have realized that beyond direct operational margins, the benefits of gained efficiencies in driving behavior, improved fuel efficiency etc. adds to their profits significantly over time.

So who leads adoption and how ....
Today there are multiple entities that are attempting to jump on the telematics bandwagon. From what is available in terms of telematics based solutions in India, the top 5 OEMs control nearly 90% of the total market. However, most of these are basic telematics services - and are not complete telematics solution packages.
It is clear that it is for OEMs to lead the push for true telematics adoption in India considering that consumer interest is generally low and absence of referable business models around telematics in India. Our interactions have indicated that OEMs in India are still learning what would be feasible in the Indian market and if at all, there is interest in the market. Hence, whatever solutions are available today are small-scale mostly pure-tracking solutions. However, knowing the potential in India, OEMs are now looking at supporting real-time data flow from vehicle to back end for true telematics solutions. As the scale of the solution expected is larger, OEMs are experimenting with building platforms on collaborating with third-party partners to sell to their customers. At the same time, considering that this is early times yet, some OEMs do not want to put too much stake and generally expect partners to lead and drive solution development, infrastructure hosting, service provider mediation and solution deployment based on terms of agreement.
Telematics platforms, built by OEMs or third parties are not limited to focusing on a specific vehicle family or integration with specific services. These platforms are being developed and localized for market-specific services. OEMs like Volvo for example, have a strategy where they provide open interfaces from their solutions. Solutions for dealers and OEMs are managed in-house while for customer/driver specific revenue-making services related to logistics and transport operations, they partner with third-party providers. The key is in developing an open NGTP like platform that enables such collaboration.
Telecom companies are also sensing business potential in developing telematics platforms or acquiring them through strategic acquisitions. The acquisition of Huges Telematics Inc. by Verizon was a potential step in this direction.Telecoms are also realizing that voice and data based telecom services are leveling out and hence packaging telematics services with attractive subscription plans could be potentially new revenue models. There are also individual telematics service organizations who can work with telecom service operators to sell attractive packages to consumers.

Companies are facing a challenge in determining a business model that would appeal to potential customers in India. In a market where infrastructure is still maturing and price is a major factor, companies will be required to innovate and create value propositions at low prices. OEMs are still not sure if Indian consumers are ready yet to adopt telematics services in their vehicles as a necessity rather than a luxury.However, until that time, it is important to keep educating users of the benefits in the long run. With the potential end user base being large - ranging from vehicle owners, fleet owners, dealers, insurance companies etc., offboard telematics in India is set to take off.


Subscribe to this blog's feed

Follow us on

Blogger Profiles

Infosys on Twitter