The Infosys Labs research blog tracks trends in technology with a focus on applied research in Information and Communication Technology (ICT)

« December 2012 | Main | March 2013 »

January 31, 2013

Trends impacting future of supply chains

Supply chains of today can be characterized as being global and widespread with multiple partners or vendors involved across the globe. Supply chains have evolved from companies that manufacture most of the components in-house to one concentrating on core competencies and sourcing rest. This transition from in-house manufacturing to outsourcing happened over a period of time, predominantly enabled by changes in economic environment, evolution of technology and production methodologies. For example improvement in Information Technology enabled teams at multiple locations to collaborate on new product development, reduction in trade barriers enabled easier access to foreign markets. A massive shift in supply chains occurred due to labor intensive manufacturing moving from developed countries (with relatively higher labor costs) to countries having comparative advantage of cheap labor.

Current global supply chains are heavily influenced by the changes in technology and economic environment. If one has to predict changes in future supply chains it makes sense to look at some of the emerging trends in these areas and their influence on supply chain. This blog will showcase current environment and identify some of the economic and technology trends that will shape up the future supply chains.

Current Environment

Global financial crisis that erupted in 2008 caused a major challenge for supply chains in terms of excess capacity being available. One can claim this by using Baltic dry index as proxy for global supply chain demand. Baltic dry index tracks the price of moving the raw material through sea. A rising Baltic dry index indicates strong demand for raw material and that in turn indicates higher global growth and vice versa.  Baltic dry index at the end of December 2012 (below 700) was near the lows attained during December 2008 (around 670).
Looking ahead in 2013, global aggregate demand may not recover due to still unresolved US fiscal cliff, most likely contraction in Euro zone, tensions in Middle East and slowing economic growth in China & India. This scenario of sub-par growth to a moderate contraction in global demand will translate into global supply chains being under-utilized.

Trends that will impact global supply chains in longer run

Current challenge of excess capacity requires businesses to focus on operational efficiency so as to manage stress on their balance sheets for survival. However this may make businesses miss out on broad structural changes that are taking place in the eco-system. While operational efficiency is required for keeping a firm afloat in short term, structural changes happening in environment may question existence of supply chain models of various organizations in the long run.
To predict possible structural changes in global supply chain we need to look at the trends that are emerging. Below are some of the mega trends that will shape supply chains of future:


Possible structural change

Rising labor costs in countries that currently provide cheap labor

USA regaining its manufacturing strength in certain sectors

Large scale shale gas discoveries in US and improvement in fracking technology leading to lower energy costs causing a few industry sectors in US to become globally competitive again.

Weakning US Dollar compared to other currencies.

Additive manufacturing or 3D printing: This involves creating a 3D object by adding successive layer of material.

From holding finished goods to raw material

Current supply chains have been built for forward logistics i.e. from raw material to retailer. An increased focus on sustainability or steep increase in commodity price will require organizations to manage reverse supply chain i.e. from retailer to manufacturer for final disposal.

A new eco-system for handling reverse logistics in efficient way

A few of the trends mentioned above have been in existence for a while, so a question that comes naturally is why things will change in future when they have not changed yet? An answer is these trends are still evolving and they haven't reached a critical level yet. My upcoming blogs will elaborate on these mega trends and their possible influence on future of global supply chains.

January 23, 2013

Why don't Bees Teleconference while Building a HIVE?

Self Organization in Teams-Learnings from Nature

J. Srinivas, Shilpi Jain, Sitangshu Supakar

SO_1.jpgWhat do pack of wolves, pride of lionesses, bees and ants have in common. What can we learn from them? What is self-organization (SO) and how does it form?

We are exploring different ways to induce this behavioral skill in the team members for greater commitment, motivation and accountability to the work. Many of us think, what is so great about it; we are self-organized and perform our daily course without fail. But, the question is can we perform equally well in a project, during crisis or with reduced resources.

NATURE has tuned the self-organized system. Be it the conduct of animals, insects, or eco-system, nature organizes optimally. What are the attributes of self-organization derived from the nature?  Can project teams organize themselves, the way nature does? Is it meaningful to compare the dynamism of NATURE with the dynamism that organizational teams face?

Before finding answers, let's understand with few examples how self-organization is an adaptive attribute in animals and insects. Imagine how the pack of animals like wolves and lionesses hunt? How honey bees organize their affairs so well in their hive and devote themselves to the welfare and survival of their colony?

Wolves are known for their intelligence and social behavior. They organizeSO_2.jpg themselves for the hunt and care of their group. The motive of the pack is to be as successful as possible, no matter if they are not the strongest one. The whole objective is to make their hunt a success so that every member can get the sufficient food. Each wolf in the pack plays a role. There is always a leader in the herd (pack) but while hunting, it rarely interferes or directs its fellow animals (Michael, Wolf., 1995-2005). Another interesting thing about them is their sense of communication; they follow communication protocol and communicate in many ways (body language, gesture, and expression). The selection of communication mean is highly dependent on the distance between the two wolves. If they are close to each other the communication is non-vocal. Similarly when they are in large group, they do 'Mob-greetings'.

They share a common objective - food for the pack. They have communication protocols and established patterns for hunting, individuals know how to respond to change to meet the objective. Their play mirrors the hunt patterns.

Let's see how bees organize themselves and find the flower nectar. Bees are deaf hence they perform a series of movements called as 'waggle dance'. These dancing steps help to identify the source of nectar and also teach other workers about the location of food source 150 meters away from the hive. The bees have orchestrated movements for communication. Especially when they are hunting for flower nectar, the experienced bees walks straight ahead, vigorously shaking its abdomen and producing a buzzing sound with the beat of its wings (Debbie, 2011). The distance and speed of this movement communicates the distance of the food site to the other bees. Another exciting aspect is the group size, the bees' colony size varies from 20000 to 80000 worker bees and they all work in coordination with each other without much direction and guidance.

The above examples of honey bees too display those benefits of the self-organization concepts discussed above. Adherence to shared objective set of practices, pattern of behavior, and communication. They show the benefits of self-organization, i.e. commitment, efficiency, and achieving self-sufficiency for the community.  Members of the community organize themselves repeatedly and continuously to meet changing requirements.

In a Direct communications with partners, iterative processes helps control conflicting interests and help them to adapt quickly to unpredictable and rapidly changing environments (Monterio et al., 2011).

  • In a research conducted by Hoda et al. (2011), it was proved that "balancing freedom and responsibility, balancing cross-functionality and specialization, balancing continuous learning and iteration pressure uphold the fundamental conditions of self-organization at certain level."

Agile manifesto stresses on self-organizing teams, and we explored what techniques make the teams achieve a sense of teamness and spontaneous adaptability which makes it work in short sprints and what will make it work in the long run. In subsequent blogs we will learn how the concepts of self-organization can be brought in a structured manner and help teams adapt in a changing environment. The resulting framework would help us in recognizing when SO can be formed or in creating the right environment for it.

Our goal is to deconstruct the key concepts of the above examples and apply them in real teams to make it spontaneous and easy to transform into a self-organizing team. Support for the concepts comes from couple of papers we looked at.


Cao, L., & Ramesh, B. (2007). Agile software development: ad hoc practices or sound principles",. IEEE Computer Society.

Debbie, H. (2011). Honey Bees - Communication Within the Honey Bee Colony. Retrieved September 13, 2012, from

Hamdan, K., & Apeldoorn. (1989). How Do Bees Make Honey? Retrieved September 4, 2012, from A. Countryrubes Web site:

Hoda, R., Noble, J., & Marshal, S. (2011). Developing a grounded theory to explain the practices. (S. S. Media, Ed.) Empirical Software Engineering.

Karhatsu, H., Ikonen, M., Kettunen, P., Fagerholm, F., & Abrahamsson, P. (2010). Building blocks for self-organizing software development teams a framework model and empirical pilot study. International Conference on Software Technology and Engineering (ICSTE), (pp. 297-304). Helsinki, Finland.

Michael, Wolf. (1995-2005). What are Wolves. Retrieved September 4, 2012, from Wolf Ranch Foundation:

Monteiro, C. V., da Silva, F. Q., dos Santos, I. R., Felipe, F., Cardozo, E. S., Andre, R. G., et al. (2011). A qualitative study of the determinants of self-managing team effectiveness in a scrum team. Proceedings of the 4th International Workshop on Cooperative and Human Aspects of Software Engineering (pp. 16-23 ). Communications of ACM.

[1] The image of wolves hunting is taken from the source:

[2] The image 'bees at work' is taken from the source:

January 21, 2013

Automotive/ Mobility/ Social

Couple of weeks ago my car started having problems with its steering wheel, on inspection by the un-friendly service manager, I was told that the software needs an update. I was also informed that there has been a recall of fuel pumps for my batch of cars, information that the car company failed to convey in time. This got me wondering, why can't there be an option where the software gets a direct download into the car as and when an update is available. Why can't there be an option where if there is a parts recall, the message gets flashed directly in the car, based on the make, year and batch of the car.   

A quick search for such options reveled multiple products available in the market (Europe/ US), take for example, the Delphi Car Connectivity Service. This basically provides for remote access through a cloud system that connects your vehicle to your smartphone by which you can remotely monitor your vehicle's overall health, performance issues and track driving behavior at any given point in time. Imagine this, based on your vehicle's performance, you can practically predict when you are likely to need a part replacement or an engine oil top-up. You can set various performance alerts for your vehicle and have access to the nearest service center all on your smartphone. At a concept level, what Delphi has done is - combined context (vehicle), remote access (location based) and mobile (anytime access) to come up with a diagnostic service, that may well be the answer to vehicles becoming smarter (on their own).

To extend the above logic, lets for a moment, assume that a vehicle could talk to another vehicle or to a physical space or even a person. Essentially paving the way for a platform meant to undertake multiple activities performed by your vehicle, without your interference. For example, can your friend get alerts the minute you are closer to where you were supposed to meet him/her. Can your vehicle alert the restaurants that you intend to visit, even before you approach the place, who in turn may end up making you offers/ deals. Well, all this is already possible given that products like Garmin Streetpilot, with Foursquare and Glympse integration are already available in the market, essentially bringing in location services, loyalty management and social media on one platform anchored to a vehicle. Looking at the concept of smart mobility in the automotive landscape, one can be assured of some very interesting concepts to be introduced in the near future and some that we can expect to become industry standard.  

January 14, 2013

DynaTrace - Application Performance Management Solution

DynaTrace Software is a leading application performance management tool and is being widely used.It comes with advanced features for monitoring Java and .NET applications, which aids to identify bottlenecks or errors in the application easily.PurePath technology used in DynaTrace provides end-to-end transaction level details; from browser, across all tiers and database.It helps to uncover performance issues even at the code level and also details of transactions invoking external services.This tool detects abnormalities in response time, transaction rate, throughput and system usage.


DynaTrace introduced to performance testing and performance engineering has helped to diagnose and fix many performance issues at an early stage.Its ability to dig deep even to code level aids in root cause analysis of the issue.

For example, provided below is the purepath  snapshot of a transaction which took high response time. i.e. > 93 seconds.[SLA : 2 seconds]Just a click on the transaction name and DynaTrace drills down to the exact web service operation, checkoutItem and displays the exact child method which consumes time.




DynaTrace helps in optimizing the performance of web, non-web, mobile, streaming and cloud applications.It supports VMWare and EC2 based clouds. This can be integrated with major testing tools like LoadRunner and SilkPerformer. Dashboards which can be customized according to the requirement is another feature of this tool that aids reporting.

January 3, 2013

The Retail Shrink Problem

The next time you are in a retail store and you hear a loud beep at the exit, only to see somebody being stopped and their bags/ bills being checked, you can be sure that the retail outlet is relying on some form of 'anti shrink' solution. You would be surprised to know that retailers in the US, lost about $41 Billion (source: Global Retail Theft Barometer 2011) last year to some form of shrinkage or the other. So what is this shrinkage problem all about? Simply put, it's the discrepancy between physical inventory and the recorded inventory through its movement in the supply chain (from the manufacturer through Point of Sale). Shrink can be attributed to multiple reasons, major reason being theft by employees and shoppers (shoplifting), and in a majority of instances, theft is committed in connivance with each other. According to the 'Global Retail Theft Barometer 2011', the main causes for 'shrinkage' was Shoplifting (35.8%), Employee Theft (44.1%), Suppliers/ vendors (4.2%) and internal error (15.9%).

Almost all the retailers have been focusing on 'theft management', using multiple solutions, these include - Electronic Article Surveillance (EAS) systems, RFID tagging, Point of Sale (POS) exceptions tracking solutions , video surveillance etc. Other solutions include - hiring people and training them on theft prevention, routine audits to track exceptions between various points of entry and exit in the sales process, physical security of high value SKUs, pre-employment checks etc. Despite these measures the shrink rate is still a cause of concern, mainly because these reasons

      - Solutions are designed to address symptoms (theft prevention) and are not comprehensive solutions focused on integrity of the inventory

- Reactive and event based solutions - shrinkage is recognized post the 'shrink event', and not in real-time or as a pre-emptive measure

- Data of Physical inventory is based on non-periodic physical counts, which are not helpful in any real-time visibility of inventory, thereby making any predictive analysis impossible

- Key decisions for loss prevention are based on analysis of historic data and there are no data points available to analyze the effect of any remedial measures implemented

- Shrink management is largely dependent on the people stationed on the retail shopfloor to spot potential shrink events

- Unclear ROI matrix for investment in new technologies for managing 'shrink'

I believe we need to address the problem of shrink with an integrated approach, which looks at bringing integrity to the physical inventory cycle, rather than focusing only on theft. This is possible by increasing the number of data points within this cycle to analyze, especially between the entry and exit of inventory from the retail shopfloor. A potential solution is envisaged as below,

- Comprehensive solution which gathers data at multiple points in real-time

- Avanced Electronic Article Surveillance (EAS)/ RFID tags at an SKU level

- Smart shelf and Smart cart (Infosys products) solutions which tracks inventory on retail shelf and its movement thereafter in real-time

-  Image recognition/ video analytics at POS terminal, to address potential employee/ shopper connivance, matched with POS data

- EAS/ RFID sensors at exit matched with POS data

-  An analytics engine which uses the above data to provide exception management in real-time

- Predictive/ Preventive analysis based on patterns/ trends that will enable management to 'heat map' incidence of shrink events, which could be at a SKU level, store level, geography, time of the year and other significant metrics