Casino, Gaming & Hospitality
Casino Marketers must be aware of the profound changes in the way customers research and buy their services. Casino marketing must be realigned and spending must be adjusted to be effective with the current realities of customer decision making. Customer retention through Loyalty groups, micro segmentation of the present and future customer into more relevant and accurate groups and better allocation of the overall marketing spend are just a few of the challenges facing this industry.
TI-data is working with a few global clients in developing customer analytics systems enabling near real-time actionable insight for improving decision management. We provide data driven customer analytics solutions to help casino marketers understand the customer decision journey and receive actionable insight reaching customers at the moments that most influence their decisions.
Life Sciences - Pharma
Pharmaceutical companies have an acute need to gain customer feedback through automated means. A Listening Analytics System is like a giant, dynamic, continuous focus group. It listens to patients and physicians in their own style of communicating. It eliminates the inherent flaws of 'question' design, sample selection and peer influence that can distort the accuracy of answers and interpretations. It provides a higher level of direct, honest input. Your constituents are saying exactly what they think and feel about your brand, your company and their experiences.
A recent project engaged Luminoso, an advanced text analytics system provider, to gain an understanding of the current problem in the U.S. and the U.K with spikes in measles cases. Generally, the children who got measles were not vaccinated and children were not being vaccinated because of parent fears. Luminoso monitored discussion boards for two months learning the factors motivating parents to either delay or go ahead with vaccinating their children. Unstructured text data was extracted from blogs and group sites for analysis. Thousands of posts were interpreted into topical clusters and sentiments giving the client a broad, candid view of what parents might express through an official survey or focus group. It provided the foundation for informational initiatives by a leading vaccine provider to mitigate this problem.
The Holy Grail of digital marketing is to get a brand to go viral in a positive manner. Most campaigns that go viral only measure the number of views and sometimes loosely correlate this to increased hits or clicks on offers or the website.
The Pepsi sponsored "Jeff Gordon Test Drive" is an excellent example of analyzing large volume unstructured text data to understand viral video impact. During the first month, the Pepsi Max 3:46 clip on YouTube garnered 35 million views, 100,000 "likes", and 19,000 comments. Comments are a rich resource to understand sentiment and what was perceived during the campaign.
Speed of insight is critical. Latency is destructive when something is happening that requires action now. The Text Analytics technology (MIT spin-off Luminoso) efficiently and effectively analyzed the very large volumes of text data, transforming the flood of YouTube comments into meaningful insights. The analysis included the brands, the ideas specifically mentioned by YouTube users and cross-correlated these concepts with the different types of commentator sentiment.
National defense missions involve Decision Support Systems that minimizes the cycle time between data collection and the operational decision. The growth of input from large data sets are overwhelming analysts. Existing tools are inadequate to efficiently process and analyze the vast amounts of data and are not keeping pace with changing countermeasures and threats. The need is to identify threats in complex, uncertain, contradictory and incomplete large data sets and to develop the necessary contextual understanding of the region and key elements.
Unstructured data analytics is the most critical and challenging problem facing the Data to Decision initiative.
TI-data is developing an automated Context Analytics system that will improve the speed and accuracy of interpretations and will handle the growing volume and complexity of the data inputs. Advanced natural language processing combined with predictive analytics is at the heart of the system. It requires additional technologies and support to function as an effective, single purpose, usable system. When large volumes of unstructured text is analyzed and interpreted it can provide valuable input to build predictive models of potential conditions of interest to be addressed. The goals is to find important relationships and targets in the extremely large data sets, reduce latency in the data analysis, while improving the accuracy of detections with fewer false alarms.
Early Detection To Avoid Failure
Petrobras - Floating Production Storage and Offloading vessel, offshore Macae, Brazil
Artificial Lift oil production requires constant adjustments to operating equipment in order to optimize production and protect equipment. The extreme location of the critical Electrical Submersible Pumping (ESP) equipment forces it to operate in an unmanned state. The inaccessibility of the ESP means that any failure on this equipment results in a month or more of production downtime. Multi-variable analysis of 16 distinct operating and condition data points in near real-time enabled the client to avoid operating states that jeopardize ESP reliability. The result was a new global record for ESP uptime was set, achieving 100% availability with no unplanned downtime, over a period of 4 years.
Mining - Fleet Condition and Production Optimization Analytics
Large mining sites are data rich environments with a multiplicity of installed but underutilized platforms in place.
Sensor networks, ERP, EAM, CMMS, SCM, SCADA, and Data Historian applications with little or no predictive analytics systems in place. There is a great potential for improving performance and reducing downtime costs with Analytics. High rates of unplanned failures, shortened asset lives, high premature failure rates, lead to excessive downtime and reduced mine site production. As an example, Haul truck downtime can often exceed 5-10% of total availability, costing mine operators millions of dollars of lost tonnage. Successful Condition-Based Maintenance (CBM) implementations require better information distilled from asset data. Accurate models utilizing all available operational and Tribology data can predict optimum values. When compared to actual data, Early Detection of potential causes leading to failure can be mitigated and reduce downtime.
US Navy - Destroyer class vessels
TI-data is developing Artificial Intelligence-based predictive modeling systems for the U.S. Navy, and Navies in other parts of the world. The purpose is to improve Fleet readiness, eliminate catastrophic failures, provide faster response time, reduce manpower requirements and unnecessary maintenance. Programs are being developed to support the Reliability-Centered Maintenance initiative and to implement the migration to Condition-Based Maintenance systems by leveraging the investment in existing data infrastructures. Predictive analytics and diagnostic models will provide Early Cause Detection to mitigate and avoid failures.
Real-time outputs from all operating and condition data points are generated from each ship. Analysis and diagnosis of machine condition requires significant man-hours from a limited pool of shore side technicians. This "Big Data" environment has overwhelmed technicians, resulting in a 'data analysis to action' time-frame of 6+ weeks. An automated system that analyzes all data, builds predictive models of "normal" machine behavior and machine conditions, compares the model to the actual data and automatically detects "abnormal" machine conditions that require attention provides timely insight to mitigate potential critical failures.
Department of State - Argentina, Brazil & Chile Embassies
TI-data built and managed the data handling system, including special encryption devices, between the U.S. Department of State communications center and the U.S. Embassy locations in Argentina, Brazil and Chile.
U.S. Customs and Border Protection - Rotterdam, Netherlands
TI-data built and managed the data handling system supporting two U.S. Custom Houses located in Rotterdam, Netherlands with connectivity back to the U.S. communications center.
Telefonica, headquartered in Spain, was developing into the largest internet service provider in South America. TI-data provided large volume data handling access between Brazil and their European interconnection in the U.S.
American Commercial Lines
TI-data built and managed the data handling system for collecting data from their barge station in Rosario, Argentina and delivering data to their center in Jeffersonville, IN.
Latin American Projects -
In 1996, TI-data started as a provider of data handling services supporting U.S. organizations with remote offices in Latin America. We provided technology and systems integration with strong in-country resource partners certified in the leading technologies. Our projects included building infrastructure and transporting data to remote locations throughout Mexico and South America.