Today, airports, stadiums, industrial sites, retailers, and many other businesses work with big data. Working with big data can be effective and simple if you are using modern tools. The SAP HANA platform is one of such tools.
Let’s take a look at some interesting case studies. These are cases taken for medium and large businesses, but there are also consumer scenarios that may be interesting to everyone.
SAP HANA in crowded places
Airports have a huge number of systems and services integrated with each other. Every mistake made while conducting routine airport activities can cost a lot of money and lead to a loss of passenger loyalty.
Of course, SAP HANA will not be able to change the direction of the wind and stop the snowfall so that the flight takes off on time. But the SAP HANA platform can be a great helper in resolving air harbor’s problems thanks to:
- Support for transaction-oriented and analytical (OLTP and OLAP) applications without data duplication.
- Opportunities to work with different data sets (from well-structured data to abstract structure data in the form of graphs and unstructured text data).
- Tools for developing and running applications in different languages.
So, what problems may arise at the airport? The most unpleasant for everyone is the abnormal operation of systems that are not ready for heavy and peak loads:
- Failures on information boards, which cause panic among people who wait for arrivals or are late for the flight.
- Unavailability of registration systems, leading to mass gatherings of people near the registration desks. Registration agents spin like squirrels in a wheel and can do nothing. In the meantime, passenger loyalty goes down.
- Traffic jams when approaching the airport (It is better to spend time, not in a traffic jam but doing purchases in the Duty-Free, it would be more profitable for the business and much more pleasant for the passenger).
- Long queues of those waiting for passport control, which arise due to incorrect calculations\management of staff shifts, as well as due to the breakdowns of automated equipment (turnstiles, inspection frames, etc).
How can SAP HANA help here? Applications inside SAP HANA are empowered to predict peak loads, quickly receive and process large data streams, and recalculate the necessary amount of equipment and employees in real time. It is also worth noting that SAP software solutions are well adapted to real-world situations.
Let’s consider a specific case. There are 20 entry gates\barriers in the airport parking. In order to track violators, the license plate recognition camera strictly fixes the car number. It is winter, weather conditions get worse and all license plates of cars get covered with snow. In addition, by a “happy” accident, 10 barriers are out of order. As a result, we see a huge traffic jam at the entrance. But SAP weather forecasts, cameras at the airport entrance and sensors from broken barriers have already predicted the traffic jam and suggested to the responsible persons that it is necessary to send engineers in advance to repair the equipment, to put the operative group of employees on the working entrances – to help drivers and control traffic. SAP HANA applications also “recommended” to lower the license plate recognition threshold, so that the camera, after recognize at least 3 characters on license plates, would allow the car to be parked.
This is just one simple and very common story that can be prevented with data science. In addition to the airport, the same kind of problems can occur at stadiums, train stations, in large shopping malls. SAP HANA is able to operate effectively there as well.
SAP HANA Smart Data Streaming (SDS) technology helps to transmit and process data in real time. SDS receives information about events from various sources, can combine them, analyze, process and transfer to other applications for further processing or storage.
Immediate processing of newly received signals allows you to send a warning, react to changing conditions, supplement the received data, and even make conclusions whether to save the data or not. SAP solutions are suitable for working with streams of signals from devices, with online traffic as a result of clicking on web pages, with events from social networks and business systems, and other tasks.
Due to the increasing competition, retailers find it difficult to surprise buyers with frequent discounts. The consumer spends a lot of time searching for the right product and “seemingly” makes his choice, focusing on price, the convenience of delivery and service. But that this choice does not always fall on your store. Therefore, retailers are faced with the task of monitoring customer behavior, making personal offers to them, thereby stimulating a purchase right here and now. In addition, the retailer may have other goals than sales: boost traffic, raise margins, get rid of illiquid goods. And all this requires a personal approach.
Today, targeting has reached a new level and helps to decide whom, what, and most importantly, at what point to make an offer (and what channel of communication to use).
To implement such cases, applications inside SAP HANA collect statistics on customers, predict purchases, and predict the reaction of customers.
When a customer comes to the store, the facial recognition system sends a signal to the applications and they calculate how and what to offer the client and give the management and marketing this information. This data is also taken from loyalty cards.
To solve the problems of forecasting and machine learning, the SAP HANA provides the Predictive Analysis Library, as well as the Automated Predictive Library (APL) component, designed to automatically select the prediction model.
If these capabilities are not enough, then HANA can be easily integrated with R. To do this, a special procedure is created in HANA with an R script, which, together with the data and parameters, is transmitted to the R server for execution. Results are returned to HANA.
In addition, HANA has the ability to integrate with the TensorFlow machine learning library using the External Machine Learning library (EML). The SAP HANA youtube channel has video tutorials.
Complex technical installations often raise questions even among competent specialists. Such installations can occur both in production and in everyday life (for many people, the work of their car engine is a complete secret). Misunderstanding what is happening at the moment with the equipment on which you work is a big problem.
Let’s recall our childhood, many of us loved to create models of airplanes, tanks, ships. Cut, assemble and glue parts. And now let’s imagine the work of a large chemical reagent plant with difficult and expensive equipment, dangerous chemical reactions. It would be nice to be able to create an analog of the production line at such a plant in order to experiment and to look at the reaction of the equipment in certain conditions. That’s what a digital twin is for.
At the moment, the creation of digital twins makes sense for really expensive and complex technical infrastructures, which, in case of damage or incorrect work, will either destroy everything around or fall apart themselves. Also, digital twins are being introduced on large-scale production lines, for example, automotive conveyors.
How are digital twins created? The first stage is mapping when information from sensors is stored in the SAP HANA platform database, which is transmitted in real time to the processing and analysis system. If, for example, we have data for hundreds of cars in the database, it makes sense to remotely monitor their condition and understand what car and when needs the oil change, etc. Equally important is the ability to use information about a particular car from our imaginary fleet to understand how it will behave in a particular “live” scenario. It is even better if the information comes to us from hundreds of identical cars, and we (having enough statistics) and experimenting with digital twins, understand that at 120 km/h the brakes may fail or something like that Thus, we already have the opportunity to build an automatic model. And such a scenario helps not only to develop your production lines and warn about possible incidents but also to save money.
The SAP HANA platform is already quite effectively working with big data and machine learning, solving the real problems of both ordinary people and big companies. The process of deploying new electronic systems and technologies within an enterprise should co-occur with hardening the organization’s overall security posture. This ensures that the new workflows are hassle-free and aren’t susceptible to malicious interference. Enforce effective password policy, maintain data backups, and conduct phishing awareness training of the personnel. When it comes to software practices, use reliable antivirus solutions, effective VPN tools, and make sure all applications are regularly updated.