Data Analytic approach to Aviation: The Sri Lankan Context
 |
| New inclusion to Sri Lankan airlines fleet : the A320 neo, a much greener aircraft |
Despite the fact of a booming tourism industry in the post-conflict era, Sri Lanka has not yet seems to have exploited the strategic movements having reciprocal effect on the economic development in the country. Concerning aviation, as one of the dominant factor in the said domain, Sri Lanka have handled 2,050,332 passengers in total at BIA in 2016, which also recorded 604,953 passengers, a 3.4% growth by month of March in 2017, with respect to the preceding years statistic by the same month according to CAA. Although, BIA had been ranked among the Top 10 worst airports in Asia in 2015 with MRIA been named the ‘World’s emptiest airport’ by Forbes. Reports, pronounce lack of hygiene, habitual corruption among airport staff had been a major concern at BIA, although lack of connecting flights and long transit times have rather driven more passengers away from BIA to major airports in the region such as Chennai and Mumbai. Therefore, basic strategic and policy planning in Sri Lanka had been demanding a suitable framework for both complement and supplementary services in terms of infrastructure, systems, knowledge in their respective areas, being ‘Aviation’ in the context of this article. A typical example being MRIA which lacked supplementary infrastructure, even though it was to cater the excess demand for BIA been projected for the next decade. These could be misleading considering the lack of effort in developing a customer based model using modern knowledge and techniques which could predict passenger movements and provide market intelligence using the enormous data being generated in the Data driven era of technology.
Considering the statistics and expert knowledge in the field of tourism, it is eminent that tourists whom had been predicted to would visit the country have diverted themselves to alternate tourist destinations in the region, specifically to Maldives and Thailand due to the security concerns from the civil war and disaster in 2004 tsunami. As been noted above, passenger movement analysis suggests high frequency of passengers, preferring air routes around the country through alternate connecting airports. Therefore, airline passengers including tourists, are clustered ‘around’ the region of Sri Lanka, but not arriving in the country due to various reasons, some been suggested above. According to statistics, despite Sri Lanka handling approximately 52,000 scheduled and 3,400 unscheduled flights annually, these had been predominantly tourists ‘specifically’ flying in and out of the country for holiday. In addition, 20,000 flights fly across the airspace with over 8 million passengers on board. Therefore, potential revenue streams and massive amounts of data had existed around Sri Lanka, in a way as if the country is being in a glass sphere, not being able to grab the opportunity even though they are clearly visible.
Exploiting large amounts of data
The said massive number of passengers would potentially generate TBs of data, only through social network platforms daily in an hour in addition to the large amounts of operational data, flight data and maintenance records from which only a mere portion the data set is being utilized by airline operators, governments, airports etc. Hence, with the emergence of data intensive science as the forth scientific paradigm, usage of data intensive applications particularly in industrial context for customer relations, market analysis has been crucial in order to enhance or develop competitiveness in the market which the aviation industry have lacked to cope with, especially considering the regional airlines including Sri Lankan air lines.
Big Data
 |
| Real time Aviation Big data platform |
Big data is a term commonly being used to refer large volumes of diverse and sophisticated data, also mistaken to be referred strictly with Web technologies. In fact, big data has various definitions and explanations. Doug Laney (2011) explained big data using 3V: Volume, Variety (structured, semi-structured and unstructured data originating from different sources of heterogeneous nature) and Velocity (continuous high speed data streams). This is also associated with another V, often relating to value, virtual or veracity (uncertain and unverified nature of data).
The collective features of big data, having high diversity demands high computation intensive state-of-the-art storage and processing mechanisms and hence, traditional off-the-shelf techniques such as relational database management systems are not able to manage the operational complexity of the process. For example, in most proposed architectures, including that of Sujie Li et.al in “Civil Aircraft Big Data Platform” (2017), a hybrid storage module adopting “virtualized storage” in order to facilitate the variety of data in order to overcome what is known as the “bottle neck of aviation applications” that arise as a result of independent and heterogonous nature of data (Data in aviation are generated from different sources such as aircrafts, air traffic controllers, maintenance centres, airports, operators, passengers using different equipment of different manufacturers which are developed on different protocols and formats).
General Aviation Platforms
Although, manufacturers have shown great interest in the modern era with research being undertaken to adopt modern trends in technology such as Big Data and analytics especially to exploit this large, heterogeneous data of high velocity generated by avionics, flight recorders and sensors in order to improve efficiency, eco friendliness, cost reduction and resource optimisation in the forms of “Predictive Maintenance” and “Aircraft Health Management”. This in fact has had a duplex effect on both the manufacturers and operators in a favourable manner. In the study by Sujie Li et.al (2017) proposed civil air craft big data platform with a three tier architecture that integrates different operational and maintenance aspects of aviation using real time and archived data sources, based on similar work been performed by the industrial researchers. General Electrics, proposed the concept of “Industrial Internet of Things” (IIoT) as the infrastructure for their cloud based big data platform “Predix” for analysing over 35,00 of their engines along with a “Fuel forecasting model” enabling abnormal alarming and fuel savings for customers. GE has recently introduced the GEnx series of engines, the fully functional engines utilizing Big Data for real time performance optimisation and predictive fault diagnosis. Of similar nature, Pratt and Whitney deployed the Advance Diagnostics and Engine Maintenance (ADEM) tool, with the Geared Turbofan (GTF) engine, opted to manage real time aircraft performance data, transmitted to ground 24 hours a day, providing the ability to constantly monitor and observe operational patterns and maintenance issues early. In similar instances, big data platforms have been comprehensive in uplifting safety and essentially reduce costs and improve fuel efficiency by 10-15%.
 |
| General Aviation Big data feedback loop |
Similar research by Dr. Tulinda Larsen in the paper “Cross platform aviation analytics using Big data methods” (2013), Anuj Singh and Akhil Kaushik on “Knowledge based retrieval scheme form big data for aviation industry” (2015) and Jin Chen et.al on “A Big Data Analysis and Application Platform for Civil Aircraft Health Management” (2016) provides the generic framework in terms for effectively handling the sophisticated big data as the platforms having very high potential that could benefit a larger group of stakeholders of aviation industry both directly and indirectly.
Using Social media as an insight for Airlines and Airports
Social media has been a dominant media in sharing and as well retrieving information on the market, the latter being known as “Business Intelligence”: the process of using large cooperate data in order to generate useful insight. Some airline companies have adopted this technique in order to enhance customer relations, such in the case of KLM’s “Meet and Sit” service which integrates Facebook and LinkedIn profiles to the flight booking system, enabling passengers network prior to the flight. Hence, in such a way, companies could rationalize their resources to provide better customer service and therefore retention could effectively result in development of the airliner. Hence, in an industry where statistically passengers’ consult 10 airlines ticket offers to purchase 1, social media has a significant effect on the purchase decision to which the airliners should pay considerable attention.
In a comprehensive research by Sien Chen et.al (2016) based on the data acquired from Sina Wiebo, China’s largest social network platform, on passengers of China Southern Airlines researchers were able to develop a model on customer social media value and perform semantic analysis on the newsfeed. The model could be applied on mapping passenger networks on social media: passengers’ relationship with airliners and other passengers to identify the association with events in the industry with respect their influence; and identify the customers’ perspective on the airliner. With customer insight, product differentiation and mass customization for airliners are highly viable providing customer oriented services. For example, as mentioned by the researchers, in case of delays, highly valued customers could be given priority of take-off, ensuring high-quality and high-value passenger service, in fact, a highly valuable insight in order to avoid incidents such as in “Delta airlines”.
Suggestive Use Cases in Sri Lanka
Sri Lankan airlines and BIA has been operating under heavy losses or low profits for a considerable span of time but plans where been put across to develop Sri Lanka as a hub as discussed above. Data analytics could be used as one of the supplementary tool, which especially was lacking in management decision support. Different applications in the general context of aviation has been discussed in this article and the national carrier Sri Lankan could especially be benefited by “predictive maintenance” and similar technologies, given that the airliner is operating with CFM LEAP-X1A26 GE co-manufactured engines with the introduction of A320neo and A330-300s, in order to reduce ground time for maintenance, avoid unscheduled repairs and most importantly reduce fuel costs.
Tourism in Sri Lanka has accounted more Indian and Chinese tourists, 356,729 (17.4%) and 271,577(13.2%) respectively in number, a significant change compared to the dominancy by Europeans in the golden era of Sri Lankan tourism in the past. Therefore, adopting sematic analysis techniques to analyse the more lucrative European and North American markets, with a recommendation feed through social media platforms could potentially have supplementary favourable effect on the independent campaigns been undertaken by the Tourism promotions board, hence avoiding hiring expensive Promotion consultants for which sometimes results ineffective. It should be noted, Sri Lankan airlines should be capitalizing on providing competitive offerings to the customers especially in a period conceiving losses by devising passenger oriented services, based on market intelligence on regular outbound destinations from Europe and North America.
In addition, Sri Lankan could potentially identify the value added features for the airliner with passenger sentiments in social media, allowing community detection and influence propagation through the passenger networks discussed before.
Considering the areas that have been covered by other researches, determining the lucrativeness of air routes and planning airport operations using big data has had less work been performed. Despite commercial services such as OAG Analytics Connection Analyser which provide evaluations on air routes based on current air traffic information, a comprehensive connection analyser could effectively be utilized for airports such as BIA. Therefore, it is proposed, in using Social media statuses generated by passengers, especially “travelling” statuses in Facebook with an appropriate Prediction model for air route passenger flow, such as being suggested by Liu Xia et.al (2016) using Grey prediction model could essentially generate knowledge how to concentrate or at least deviate a significant portion of the passengers to use BIA as a hub. In simple words, by identifying the alternative routes taken by passengers whom could have been catered by using BIA as the transit hub, and evaluating profitability of frequent destinations through those alternative routes, management decisions on route management and device competitive offerings to uplift value or the profitability air routes arriving at BIA could be made. Hence, serving the purpose of transforming BIA as the transport hub in Asia.
In conclusion, it is pertinent to state that introducing Data analytics techniques or at least taking services for civil aviation purposes will open up new dimensions for air operations in Sri Lanka while augmenting the potential for economic development and trade capability of the island nation as Transport hub in Asia. As well providing customer oriented airline services could improve the profitability of the national carrier.
Comments
Post a Comment