Recently, I came across an article which mentions BIG DATA 1.0 was all about storing volumes of independent silos of data to provide competitive advantage to business. Organizations would toy with Data Mining architecture to generate meaningful content. There is much hype and hoopla surrounding the analysis of independent silos of data, but is it worth the effort? The deluge of data is difficult to store, capture and visualize. A proliferation of diverse datasets- from historical data to recordable data and from safety records to sustainability data is even difficult to manage.
What are the organizations doing with safety and sustainability data today? A dashboard generating some insight would be a simple answer- with difficulty being stakeholders requiring different reports and formats. If the dashboard is built on data stored independently without interfacing other datasets, the representation would hardly give any competitive advantage.
Integration is the key. BIG DATA 2.0 is about integration which makes data useful through contextualization and correlation. There is a paradigm shift in the perception that point-to-point integration would generate some meaningful content, where upgrading to next version is a key challenge. Point-to-point integration consumes resources, time and often yields unexpected consequences. The effective solution is to build a model which presents a more holistic way data is interpreted. An example would be records related to the employee i.e. Exposure, Incident, and Health & Safety records as well as personal information. The solution in the mentioned case would be an application wherein the personal records would be interfacing the application having Health & Safety record. The Health & Safety records would be analyzed to estimate the probable incidents in which employee might be involved directly or indirectly. The proactive risk management would thus enable the organization to minimize the incident rate.
BIG DATA 2.0 would enable the organization to rediscover value through a refined modeling process: make a hypothesis, create visual models, validate and create a new hypothesis again, resulting in enterprise-wide integrated platform. The key for safety and environmental data visualization is to collect data in standard formats, determine the datasets which are relevant to the stakeholders and to feed the data in an adaptive algorithm, thus providing competitive advantage to the business.