Data science is hot. Per IDC, big data and business analytics worldwide revenues will grow from nearly $122B in 2015 to more than $187B in 2019, an increase of more than 50% over the five-year period.
Not only is analytics a fast-growing opportunity, but it is rapidly evolving from an on premise reporting focus to predictive and prescriptive modelling of business processes in the cloud. Gartner says that by 2020, predictive and prescriptive analytics will attract 40% of enterprises’ net new investment in business intelligence and analytics. During the last year, IDC found the on-premises portion of the overall market contracted by 1.4%, while the public cloud services revenue grew 26.5%. Public cloud portion of the market now represents 17% of the market.
Data science is about using disciplines such as statistics, machine learning, data mining and predictive analytics to extract knowledge about the world from data in both structured and unstructured form. In some cases, methods that can scale to handle “big data” are of interest but it is possible to learn from any size of dataset. Simply having large volumes of data does not mean that you will be able to understand, analyse and learn from what you have.
SAP has recognized the importance of data science and big data and has incorporates analytics into the core of their products. For example, SAP S/4 HANA features Embedded Analytics that enable real-time operational reporting and analysis as a core content component of the delivered product. Further, SAP has released new products like Predictive Maintenance and Service (PdMS) that are exclusively focused on delivering on the promise of data science in specific verticals.
In addition to product development and enhancements, SAP is embracing the leading open-source technologies. One interesting recent acquisition in this area is Altiscale. Altiscale offers Hadoop-as-a-service which is a key technology for big data applications as it allows the storage and analysis of huge volumes of data. Altiscale/SAP takes care of the data management allowing customers to focus on implementing their use cases. Altiscale services reduce the cost of bootstrapping these sorts of projects and allow them to scale as required. These services, combined with the internally developed SAP Vora product that integrates HANA data with Hadoop, yields interesting potential for enterprise analytics in the SAP landscape. For example, real-time IoT data stored in Hadoop is integrated with SAP ERP information in SAP S/4 HANA to develop new insights.
Vesta adds value as an enabler of holistic data-driven EAM and aims to help our customers leverage structured and unstructured data sources both within and outside of SAP to gain insights into maintenance personnel performance, MRO and other aspects of EAM. Vesta is uniquely positioned to provide such a service as experts in industry and EAM processes and how to optimize them.
For Data-driven EAM, Vesta starts with data profiling and exploration to understand existing data sources and quality. We then move on to predictive modelling with a differentiating focus of being able to explain the models and translate them into real improvements in EAM. We use SAP Predictive Analytics to do the initial data analysis and reports for customers to understand what they have and how they can benefit from using this data. We then move on to enabling predictive analysis (using the base of SAP PdMS and SAP HANA Predictive Analytics Library) to solve real EAM problems and make a difference to the various industries. Even small improvements on large numbers can represent a winning business case for most organizations
The SAP PdMS approach is shown in the diagram below:
The typical skills we provide to do a complete data-driven EAM implementation are as follows:
Domain knowledge. Without industry specific and EAM understanding, it is difficult to develop analytic models that deliver value.
Communication skills. Delivering the insights so action can be taken is vital to business success.
Analytic skills. An understanding of how to create, apply and evaluate models for the task at hand.
Technical skills. These are the skills required to retrieve and shape the data for analysis. This includes programming skills, database skills and an understanding of how to work with unstructured data.
Intellectual curiosity. A passion for solving problems in novel ways.
At Vesta, we leverage our EAM expertise to translate analytics into real results enabling our customers to succeed with data-driven EAM.
Ever been involved in a major systems replacement or change program?
I have! Memorable journeys all of them. The initial excitement of the challenge, the frustrations of the build, the stress of the cut over, and the satisfaction of a successful go-live. Then the fun really started! Remember all of those benefits? You may not have promised them, but there is a fair chance that you are going to be expected to deliver some at least; and in my experience, asking management to hold its collective breath for a year, while it all sorts itself out, just doesn’t cut it. You need managements continued support, but how do you get it? You are going to need something tangible and that’s where the reporting and metrics come in.
It’s not surprising that a business needs to know how it is performing. What is surprising is that many businesses don’t do a very good job of identifying the right things to measure and putting in place the appropriate business processes to ensure what they do measure, is accurate and relevant.
You may not even start with a clean slate. More often than not, most of your initial data is going to simply be what was available before. If it was good you’re in luck, if not, well now you just have another problem because all the good work you have done may be constrained by that same poor quality data.
While we might not like it, we know that measurement is essential. You need it to determine goals, establish performance, observe trends and track improvements. So once you have decided the things you are going to rely on to objectively state performance facts, (the kpi’s), you need to have a good look at the data you are using to assess them. Data is often overlooked and not recognised as the “business asset” it really is. Implementing a set of metrics around the data, (measuring the measurements) is not such a silly idea at all. In fact if you aren’t doing it, there is a very real chance those kpi’s are going to become very beastly, as will the managers that might be accountable for them. Developing a suite of metrics around data that check its quality (typically accuracy, completeness, relevance, currency etc.) to either make sure it is good, or if not, drive business processes changes to fix it, actually makes a great deal of sense. Once we have confidence in our data, reporting become easier and kpi’s can be relied upon.
There is a whole other story about lagging or leading kpi’s and whether they measure process or outcomes. Rather than wake the beast again we will leave that discussion for another day.