Big data in Supply Chain Management

Big data in Supply Chain Management

Abstract:

Globalization is both a boon and a bane to businesses. Businesses can now have a global presence and are no more geographically restricted. At the same time, they have to compete hard against global players in the market. Quicker and better decisions help businesses gain an edge over their competitors. Big Data gives a doozy opportunity to make better and informed decisions in many facets of business let it be marketing or manufacturing or anything else. Supply Chain Management is an area of business that is deeply impacted by Big Data boom. Excellence in the supply chain is a definite, competitive advantage for businesses and Big Data can be harnessed for a more efficient supply chain management, even under volatile situations. This essay presents the scope of Big Data in Supply chain management, the online and offline data processing and various Big Data applications. The desired outcomes of Big Data and its business and organizational impact are presented before drawing conclusions.

 

Introduction:

Big Data refers to voluminous data ranging from a few terabytes to multiple petabytes and is highly unstructured and may contain a wide variety of information such as text, images or audio/video, time series data or spatial data or other forms. The volume and variety of Big Data often require new algorithms and architecture to manage and extract knowledge from it.

Further, the rates at which data is generated, i.e. the velocity of data, is another factor which distinguishes Big Data. There might be incomplete information, ambiguous information leading to inconsistencies posting a question on the veracity of the data. These characteristics of Big Data are called as the Four V’s of Big Data.

Database management systems were used traditionally by organizations to process real-time transactions, and the technology is called OnLine Transaction Processing or OLTP. The data warehousing technology allowed OnLine Analytical Processing or OLAP. But data today is generated by each, and every one of us and hence the volume is too huge for such systems to handle. The term Big Data encompasses the technologies and initiatives that involve data that is too unstructured, too complex and too massive for conventional systems to handle. Such technology can carry out Real Time Analytic Processing or RTAP.

Supply Chain management is very crucial for any organization which is globally connected. It is an important an important area where a variety of resources have to be managed. Raw materials, transport, and logistics, seasonal demand and availability statistics, water availability, climatic conditions, customer requirements, etc. form a subset of the information that must be processed for creating an effective supply chain. RTAP or Big data is often used for effective supply chain management.

 

Online and Offline Data:

Big Data is created online, and analysis can be done in real time too. Often the data is stored and is used for analytics. The former is called online Big Data processing and latter is offline Big Data. The context, nature of the query, the time available, the kind of IT infrastructure used, etc. are factors which decide whether online or offline Big Data is used. We will look at two situations in the supply chain context.

Some decisions such as the choice of a raw material supplier have to be in real time. Further, it is a difficult problem because of a whole array of options such as cost, availability, delay in shipping, transportation costs, duration of transportation, etc. have to be considered before making a choice. Also, the performance of the vendor, the quality of goods he has sent in the past, installment options, discounts, etc. might also be taken into consideration. Having so many different factors affecting a single decision, data analytic strategies are used, but the solution has to be available in real time. This scenario is an example of online data mining.

On the other hand, some decisions, such as the seasonal behavior of data require data observed over a wider time span, and the analytics too could take a lot of time. If the decision is made in longer time, then the analytics can be carried out separately on already stored data. An advantage of such an effort is that it can consider a humongous amount of data, but the process is time-consuming, and the result is not computed in real time. The decisions made on the seasonal offers that could be provided is typically an example of this type of analysis. The processing for these questions is usually done offline and hence called offline mining.

 

Technologies Available in Big Data Solutions:

Various applications are available to store, process and carry out online and offline data analytics on Big Data. In fact, for every industry and organization, in every domain, Big Data can help if the right application is used. It is very fundamental that the right Big Data technology is chosen for effective Analytics while not incurring too much expense on the company.

The Big Data landscape as shown below has too many entities as shown below:

 

The Hadoop, MapReduce, Apache HBASE and Cassandra applications are specialized file systems.

In this section, we would very briefly describe different Big Data technologies available today. This section will prove to be rather useful in deciding the right Big Data technology for our project or business going forward.

Hadoop

Hadoop is an open-source software for the purpose of processing Big Data across multiple parallel servers.

MapReduce

This is an architectural framework on which Hadoop is based upon.

Scripting Language

Scripting Languages are programming languages that works well with big data.

Machine Learning

Any software which is used for the purpose of finding the best possible model that fits the data set under consideration is known as a machine learning software.

Visual Analytics

The tools which specialize in displaying analytical results in visual and graphical forms come under this category of Big Data technology named visual analytics.

Natural Language Processing

This is another technology for analyzing Big Data. The set of players which fall under this one is the software’s which can be used for analyzing different forms of textual data.

In-memory analytics

This Big Data technology of In-memory Analytics is used for processing big data in computer memory for a massive boost in computing speed.

 

Selecting Big Data Applications and Desired Outcome:

Though there are many Big Data technologies available out there in the market at the moment choosing the one which best suits the needs of our business or project becomes important and critical too. The primary reason for this being so important is that there is no point spending money on technology which won’t add value to the organization or business substantially more than what we would be investing up on it. At the same time these are cutting edge technologies, and hence the organization or the business owners need to invest time as well as money either on existing human resource available or hire new ones with these capabilities just to make use of Big Data technologies the project owner just invested in. Hence we need to be smart enough to invest only in the right technologies which will add value to the business and the project. Second important reason to invest in the right technology is to make sure that we don’t get overwhelmed with the data at our disposal and start analyzing each and every step which finally won’t be of any use to make business decisions.

Coming to deciding up on the technology we need to use in our business we first of all need to understand that Big Data varies with business application. This also means that the code used to rearrange and process data as per our requirements also varies. As here we are considering that the business in which we are applying Big Data technology is for supply chain management we would try and get hold of the best Big Data technology that we need to take further for investment according to the business’s perspective. Hadoop is a Big Data technology which uses a framework for processing the data namely MapReduce for distributing data across disks as well as for challenging computations. Moreover, MapReduce commands are processed across many nodes in parallel on a big data platform followed by assembling the same as the desired output. These are few things which could prove to be extremely beneficial for our supply chain management business which on most occasions would require to make multiple computations at the same time and throw out the results quickly. Apart from this supply chain management is a subject which is more easily understood and explained across teams and higher management through different graphics rather than sheer numbers or few basic descriptive statistics. This could be taken care of in Big Data environment with the help of its sophisticated visualization tools. This is all we should need to start basic applications of Big Data in a supply chain management business and hence finalize on these 3 Big Data technologies namely Hadoop, MapReduce and Visual Analytics.

 

Business Impact of Big Data:

The impact that Big Data can have on any business, in general, is expected by most industry experts to be way too much. Considering the business under our study, the business of supply chain management Big Data will surely play an extremely important role here too. Big Data could well transform the business of supply chain management. The changes could be made at most steps of the process to extract maximum value from the business. These could well include processes like making the Big Data technologies help us decide the optimum path to be used in the supply chain. Though few may think this to be an optimization problem with a lot of tasks (may be in tens of thousands) to be met in today’s world along with lots of constraints of might be equal order to be met this problem surely turns up to be a problem of Big Data. The two of the three Big Data technologies which we decided to apply in our problem initially (Hadoop and MapReduce) can surely help us in impacting our business by helping us in making faster manipulations of data and addressing ad-hoc requests.

 

Organizational Impact of Big Data:

The third Big Data technology (Visual Analytics) which we decided to incorporate helps in impacting the organization. Most important stuff in this direction is making the Big Data i.e. the large sets of data available with the organization in a manner which could be read and understood by a major section of the employees of the organization. This will make easy for business delegates to share their insights and vision for coming period with employees at the other levels of the organization. This will prove to be extraordinarily useful in helping to share the insights derived from Big Data with everyone in the organization including those who are not used to deal with data. In such a way Big Data technologies can impact the organization in a huge way.

 

 

 

References:

Aaron McKenna, M. H. (2010, July 12). The Genome Analysis Toolkit. A MapReduce framework for analyzing next-generation DNA sequencing data.

Gang Wanga, A. G. (2016). Big data analytics in logistics and supply chain management: Certain investigations for research and applications. International Journal of Production Economics, 98-110.

Hsinchun Chen, R. H. (2012, December). BUSINESS INTELLIGENCE AND ANALYTICS: FROM BIG DATA TO BIG IMPACT. MIS Quarterly (SPECIAL ISSUE: BUSINESS INTELLIGENCE RESEARCH), pp. 1165-1188.

J.J. Thomas, K. C. (2006). A visual analytics agenda. IEEE Computer Graphics and Applications, 10-13.

James Manyika, M. C. (2011). Big data: The next frontier for innovation, competition, and productivity.

Kunze, M. (2015). Big Data Technologies. In Big Data and High-Performance Computing (pp. 69-81). Heidelberg: IOS Press.

Matthew A. Waller, S. E. (2013). Data Science, Predictive Analytics, and Big Data: A Revolution That Will Transform Supply Chain Design and Management. Wiley Publications.

Sanjay Jha, M. J. (2014). Integrating Legacy System into Big Data Solutions: Time to make the Change. 2-8.

Thomas H. Davenport, P. B. (2012). In P. B. Thomas H. Davenport, How Big Data Is Different (pp. 43-46). Cambridge: MIT Sloan.

Vinod Kumar Vavilapalli, A. C. (2013). Apache Hadoop YARN: yet another resource negotiator. 4th annual Symposium on Cloud Computing. California: SOCC.