Miras applications includes 5 categories: airline operations, flight planning, data infrastructure-ready for analysis, market analysis and revenue analysis based on reservation data. In all cases, Miras benefits from a fast distributed data platform which eases data management handling and is able to be integrated with a variety of analytic applications developed separately by different computer languages.
The only and very traditional way of communication between airports and airlines in the world is through messages called DCS messages. DCS data fed by reservation, pre-loading and other systems by airports, includes all flight information from passenger lists and services to load and weight info. In this way, DCS data provides a rich environment for a wide range of intelligent analysis in predictive and recommendation systems which all leads to a passenger market-based product and increases revenue during the time. It also improves operational chores utility which leads to lower human interaction.
Capabilities mentioned above are applicable when a correct data infrastructure is available to manage bulks of inherently unstructured data which are DCS messages. Here, we benefit from ActorFS platform for data storage and retrieval and then bypass complexities of database model design which are inevitable. Miras DCS applications including a very dynamic and fast parser for pattern extraction, support all DCS message types format and then being a central data point for all agencies, then a reporting system over web that gives a general look through data and finally simple API’s which allow integration to other aviation systems are developed over ActorFS big data platforms.
Furthermore, one important way to reduce costs of flight for agencies is an appropriate assignment of aircraft to each flight and this is possible by analyzing histories of DCS data. DCS data also enhances load and weight balancing systems and help to moderately distributes loads inner/inter airports.
The main goal in aviation to survive is to monitor revenue stream and keep the current market share.
One of the major application of BSP system lies on monitoring revenue-cost balancing rate per different flights so to early detect flight legs which have economic losses and those with less profitable rate. In contrary, flights with highly profitable rates are distinguished. In all cases, current status is reported to top officials to reconsider their fleet planning policies.
Miras as a solution provider, proposes a big data engine to extract billing information necessary for analytics using BSP data in the first step and applies analytical libraries in python, Julia and R to develop analytical Leggoes such as to predict bestselling and worst selling cases and scenarios and then recommend how to maximize revenues and minimize extra costs.
One of the critical issue in aircraft load is to find a best-suite solution for load and baggage arrangement. Fuel consumption, aircraft depreciation rate and airport services have a very direct relationship with the subject. Optimal fuel consumption addresses best tuning between load, GC and arrangement of cargo in plane.
MIRAS as a data analytic company is dealing with cargo arrangement on plane with very high degree of accuracy based on real aircraft fuel consumption data and load planning information.
A big picture for every airline in the world is to extend its market share among other potential competitors. This has a direct impact on revenue growth.
Miras value proposition in Market analysis application is a way that provides insight for senior managers to not only view and track their records of passengers and revenue based on flights but allow them to achieve a higher level vision to compare their operations and market functionality with other local and global competitors and then pushes organizational policies to change their current inefficient market and operational process.
At regular interval, flight registers are checked against minor and major deficiency. This sometimes requires to change the current piece/segment. This process is very important since it deals with life of passengers.
A system to save history of sequential check is essentially required to see which part is replaced with new segment or which part is repaired. This information is principle and help to predict life time of aircraft and to track its status from beginning to end.
All these applications are possible when a hierarchical model to extract information of all segments for any type of aircraft is designed and the relation between elements is defined.
Miras develops a very easy working solution to input a series of unstructured data, process data in multi-level hierarchy model and output them to any data engine such as oracle, sql and etc so that data can be accessed by performing query.
When it comes to scaling issue in terms of fight register, Smart allocation of aircrafts to flight legs based on variety of dynamic conditions such as flight type, special event, fuel, weather and etc are inevitable. The complexity of these parameters grows as aircraft numbers increases, so sooner or later airlines face scaling issues.
Fleet assignment system should be able to calculate and suggest near-to-best solutions of possible choices through various flight permutation and aircrafts so that resources will be consumed efficiently and revenues dramatically exceeds in comparison to costs.
Miras is working on a rule based framework which is so flexible to add new rule/condition and provides restful API’s to access result. The result contains several sub optimal solutions that is the output of fleet assignment algorithm.
Organizations need tool to manage their business processes or activates. Activities are very flexible and can be modified during the time.
In this way, a very dynamic data platform is required to store, retrieve and access to different data model and be able to be stacked by variety of analytic applications.
As a big data solution provider, Miras has accomplished Hadoop compatible distributed data platform to manage business activates within the organization in the first step so that regardless of dealing with complexities of data warehouse design and database viewing model, they are able to add, update and maintain any type of data structure and query through all data in short time. Moreover, as business data grows, a distributed data engine can ease scaling issue both for querying and data analytics. This solution is offered and have been being in implementation phase.