Sunday, 5th July 2020

Calculating carbon emissions

GigaSpaces says that Groupe PSA has successfully launched its digital application to calculate in real-time, the CO2 emissions of each personalized car configuration to meet the WLTP (world harmonized light-duty vehicles test procedure) standard.

The solution, powered by GigaSpaces InsightEdge, delivers a 15-19 milliseconds query and analytics response time handling up to 95.2% of calculation requests without accessing the mainframe. Infrastructure footprint was reduced by a factor of 4-6X while scale was increased by 20X. The solution was developed and deployed in production in 12 weeks. Groupe PSA is the second largest car manufacturer in Europe, manufacturing Peugeot, Citroen and other brands, with sales of 3.5 million vehicles in 2019.

In order to comply with the WLTP standard, Groupe PSA needed to calculate CO2 emissions based on their customers’ configuration of features selected using the PSA online application. The calculations required a less than 100 millisecond response time to deliver the expected user experience along with a high level of accuracy and reliability to avoid fines that could reach up to €100 million annually. The mainframe modernization solution also needed to be cloud ready, scale elastically and ensure always-on functionality. The existing mainframe solution could not deliver on the requests per second requirement which was 15 times more than its performance capacity.

PSA, conducted a study on various technical solutions, but the competitive solutions were limited by storing data solely on a key value store which could not meet the performance requirements due to the need to frequently access the mainframe to execute complex queries and analytics.

“Our selection criteria focused on scalability, performance, the ability to store and process various data structures, overall simplicity and cost. GigaSpaces InsightEdge was the most suitable solution,” said Santiago Castelao, Lead Project Manager, PSA. “As we move forward, we will continue to leverage the solution’s ability to unify transactional processing with real-time analytics and machine learning and introduce new applications to meet additional business needs. “

GigaSpaces InsightEdge data and analytics platform was implemented to resolve the mainframe bottleneck problem by seamlessly offloading requests to its core in-memory data grid engine

Unlike a standard database, GigaSpaces InsightEdge uniquely co-locates data and business logic in the same memory space, which delivers unparalleled performance and reduces network costs. The cloud native in-memory computing solution optimally supports event driven advanced queries and analytics and unique aggregation and masking capabilities, which streamlines the workflow, maximizing performance while minimizing the data footprint.

“We are proud to provide Groupe PSA with most modern data solutions to deliver superior online customer experiences while complying with clean air regulations,” said Moshe Weizman, Director, France & Nordics at GigaSpaces. “The successful deployment of this solution underscores our commitment to our partners’ and customers’ success with the most advanced, agile data and analytics platform for all of their digital-transformation initiatives and legacy strategies.”

Partnership focuses on meeting demand for trust, accuracy, and traceability of data to support data...
MariaDB Community Server 10.5 turns into analytics powerhouse with cloud-native storage.
SAS’ AI and analytics more tightly integrate with Microsoft Azure; Microsoft to bring cloud-based SA...
Qlik has published a global study that shows organizations that strategically invest in creating dat...
MongoDB has introduced a series of products that comprise the MongoDB Cloud platform that give devel...
Informatica has introduced new intelligence and automation capabilities to the industry’s first Clou...
AWS machine learning and analytics to enhance the fan experience for Germany’s top football league,...
Over half (53%) of organisations also wasting time on deciding how to even collect data.