Infosys performed near real time aggregation and enrichment of parcel scans data. Early view of the expected parcel volume in depots and in fields. This was instrumental in enabling them for proactive resource planning.
Persona based dashboards: depot, client manager, field manager, and executive, parcel shop, providing real time view and highlighting underperforming nodes in the network under their control.
Volume metrics such as total volume in the network, expected vs actual volume at each node with drill down ability, carry forward volume from past, new volume added in day, number of parcels out for delivery or already delivered and many more. Ability to filter the metrics based on various parcel attributes. Drill down from org level view to parcel level.
Integration with LDAP for authentication and role based access for tight security.
Near real time visibility of expected parcel volume in the nodes enables operations and field teams for proactive resource planning and optimize resource utilization.
Real time view of Expected vs Actual performance helped in improving the service delivery. The laggards, highlighted in red, attract operations team’s attention immediately and direct their focus on clearing off bottlenecks in network parcel traffic. This was not possible earlier.
Real time and interactive view of all parcel volume metrics compared to day old reports in earlier world.
Close to 200 sensors are embedded in each of these vehicles to feed information about both the vehicle and the terrain which is sent back to a remote operations center. Each truck has 400 data points continually sending out information on things such as temperature, vibrations, and tyre pressure.
Data derived from the sensors was streamed into IIP at the rate of 27,000 messages/ second through KAFKA. A mathematical model developed in Apache Spark was applied on this data to derive the probability of equipment failure or if the equipment needs to be replaced or if the machines are in order.
Processed data is overlaid on the 'world map' – a native HTML5 application built on top of IIP. The detailed information can be drilled down by clicking on the color codes.
The IIP-based solution offered enabled data processing in real-time, and was completely elastic and scalable. This supported the client's business plan to increase their fleet of unmanned trucks by more than 300% with corresponding increase in sensor data loads.
Real-time analytics helped get a clear view on aspects like adjustments to production schedule, spare part order release, maintenance schedule, energy costs, machine availability and reliability and optimizedasset utilization. Predictive maintenance resulted in reduction in machine breakdowns with significant improvements in downstream supply chain operation efficiency and business profitability.
IIP helped predict if an ATM is functioning or may fail in a week's time (over 80% confidence level) through alerts and incident data from XMS and machine data.
Ticketing data from 8500 ATMs / 4M records were ingested into Spark and then the cleaning was carried out (date, spaces, null fields, etc.) in just 27 seconds on 10 node AWS cluster (32 CPU, 64GB RAM, 640GB SSD storage).
Enriched the data to create parameters required for analysis. Spark Logistic Regression ran on top of this data with data prediction in 60mseconds.
The data was ingested to Oracle and the visualization was done using Tableau.
There was a 14.3% efficiency increase (from 3.5 to service calls per technician per day).
Client saw an 18% cost reduction from increasing the mix to 40% staged calls.
The time taken for chronic/ defect identification came down from weeks to days and then, ours.
Analysis was done on customer data for various dimensions like transactions, product, engagement, transaction timeline, demography, service requests, and external variables for the last two revenue cycles by logistic regression to understand historical behavior in order to develop a mathematical relationship on risk factors. The churn probability was derived as a function of risk factors.
13 GB of customer data including transactions, equipment holdings and disputes ingested into HDFS. The ingested data was processed using Spark SQL and the churn probability equation was applied on the data to predict the propensity of customer churn. Tableau was used for visualizations.
The execution of 66 million records to predict churn took 6.5 minutes and was executed over 3 Amazon instances of 60GB RAM with 640GB storage, demonstrating Spark's capability to deliver exceptional performance through in-memory computing.
Significant factors affecting churn were identified and the high spending of customers among highly likely churners were prioritized for remedial strategies/promotions.
In order to eliminate unnecessary locomotive braking events IIP examined the locomotive brake data in combination with engineer characteristics, wayside data streams, maintenance zone information, environmental and atmospheric data and PTC signal data.
PTC braking event and weather data was ingested into IIP on AWS cloud. Text analytics, and a delay event prediction model was built using R and reporting/ visualizations in Tableau.
Different types of analysis were done, such as analysis of locomotives by PTC profile and attributes involved in braking events and providing a 360 degree view. Pareto analysis of Target Type of braking events, basic text mining (N-Gram)/word cloud on delay comments and locomotive delay prediction was also done.
IIP enabled real-time data collection of PTC running into 100s of TB and then real-time prediction based on this data.
IIP helped predict failures with higher accuracy in order to increase the velocity by 1MPH which can result in $200 million revenue. This also helped client's asset utilization increase from 45% and reduce the overall assets for the current demand.
IIP's Twitter Adapter was used to acquire public tweets from users across Twitter handles, to interact with the user community.
Tweets were pre-processed and stored in the IIP data lake and exposed as views/tables to data analysts and scientists. The data scientists used IIP-R and IIP-Text analytics capabilities to analyze and generate N Gram-based topic clouds, sentiments and engagement scores.
Processed data was integrated back into the IIP data lake and visualized using Tableau so that insights derived from Twitter could be communicated and shared with business users.
The IIP-based solution offered data scientists and analysts a flexible approach to source data from Twitter and possibly combine it with structured data available in the IIP data lake to generate insights related to community engagement and sentiments.
IIP analyzed and created ingestion models to ingest data from the source systems, including porting of 24 months of historical trend data for analysis. A staging and design model was created to transform and store data for payments, credits, debits, adjustments and other transaction tables. In addition,a master data for customers and collectors was also brought out from the source systems. The data model was optimized, harmonized and summarized tables in Hive were created.
IIP leveraged Spark SQL to visualize data as tableau reports. Provided ad hoc search and download interface to power users, allowing them to download large amounts of data for troubleshooting and auditing purposes. This simplified the complex reporting navigation structure, improved the report performance as well as provided export options for offline analysis.
IIP improved reporting accuracy with reduced reporting variance between old DWH system and AR balance from $30 million to just $84K. In addition, many discrepancies in discount and payment transactions were identified and fixed.
Over 400 different views for reports were reduced to approximately 50 views enabling a flexible and simplified the reporting framework. Reporting capabilities improved for liquidation of invoices over 24 months from the earlier 13-month period.
User adoption capabilities was increased with the ability to create additional reports and dashboards. The data can be combined with other sources and the marginal cost to build newer applications is significantly lesser.
Over 20 data files with finished goods data for both international and domestic sales (shipment data, sales data and product master) were ingested into the IIP data store.
Over 50 models were created into the IIP data lake to bring all the data together at item and UPC level and calculated the sales (LTM), average batch size, profit, and YOY sales increase, ACV for distribution and velocity. Threshold values were established for each of the above metrics (guardrails) to identify the SKUs that pass or fail these guardrails.
Multiple outputs were created in Tableau as a visualization tool. Interactive what-if scenarios analysis (for changing the threshold values) was also provided.
Identified a list of SKU failing most guardrails and initiate the business process to address the issues or drop the SKUs.
An automated SKU rationalization process was implemented based on IIP to identify the non-performing long tail of SKUs. This process can be repeated on demand. This enabled the client to effectively focus on winning SKUs, drive operational efficiency and procurement savings. Apart from the SKU rationalization, further reports of sales performance and predictive analytics can be developed on top of the data platform.