Reported impact
Business outcome
AETA significantly improved prediction accuracy, reduced sensor-maintenance effort, and established a repeatable AI workflow it can keep refining for broader public-warning use.
- Risk Reduction
- Scalability
CapgeminiEducationRisk ManagementAI and Machine LearningData AnalyticsRisk ReductionScalability2023 CapgeminiCAP#0070adEducationChinaAETA needed to improve the accuracy and operational readiness of its earthquake-prediction system so governments could act on its forecasts with more confidence.600 teams from 28 different countriesAug 4, 20236Risk ManagementWALKTREE prediction model · competition system · three-part sensory systems0 2023-08-04 6
AETA
Risk Management
AETA needed to improve the accuracy and operational readiness of its earthquake-prediction system so governments could act on its forecasts with more confidence.
600 teams from 28 different countries
75% weekly accuracy
88% daily accuracy
earthquakes greater or equal to 3.5 magnitudes
Solution
AETA and Capgemini's AIE Shenzhen advanced earthquake prediction by combining a sensor network, competition platform, and neural modeling workflow into one AI pipeline. AETA's three-part sensing systems collect electromagnetic and geo-acoustic data across high-seismic regions, and a competition system ingests weekly forecasts plus real outcomes so models can be compared against ground truth. The winning WALKTREE approach uses a deep neural network that first narrows likely regions through longitude and latitude prediction, then classifies expected earthquake magnitude; a lightweight network consumes maximum, minimum, and average sensor values to generate the final forecast. Alongside model improvement, AETA hardened the sensors to reduce maintenance effort, turning the project from raw data collection into a repeatable prediction workflow that can support public-warning, infrastructure, and emergency-planning decisions.
Data flow
Sensor systems gather electromagnetic and geo-acoustic data from target regions and feed it into the competition system alongside observed earthquake outcomes. The WALKTREE model trains on those inputs, predicts location and magnitude, and returns forecast results that can support evacuation, infrastructure, and emergency-response planning.
Solution architecture
- Ingest
- three-part sensory systems Collect electromagnetic and geo-acoustic signals from seismic regions before, during, and after earthquakes.
- Compute
- WALKTREE prediction model Uses a two-step deep-neural-network approach to forecast earthquake location and magnitude more accurately.
- Orchestration
- competition system Accepts weekly forecasts and real earthquake data so prediction approaches can be evaluated against outcomes.
Architecture clues
- The lightweight model uses maximum, minimum, and average absolute sensor values as inputs
- The neural approach first predicts approximate longitude and latitude before classifying magnitude
- The team improved sensor durability alongside model accuracy
- Weekly prediction rounds created continuous model evaluation against real events
Evidence from the source
"By deploying a number of three-part sensory systems in a mesh grid network across geographic regions with high seismic activity"
"By the end of the competition, the WALKTREE model had achieved an 88% daily accuracy and a 75% weekly accuracy"
"WALKTREE ... developed a deep neural network model that divides the problem into two parts."
"the WALKTREE method constructs a lightweight neural network model that takes the maximum, minimum, and average of the absolute values of electromagnetic and geo-acoustic data as inputs."
Transparency Evidence & methodology
Review how this reference was scored and trace each normalized claim back to its documented source.
Scoring methodology
How scores are calculated
Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.
Customer reference value100/100
- Problem clarity20/20Measures whether the customer's operational problem is specific enough to compare.
- Implementation depth25/25Rewards named components, data flow, integrations, and architecture detail.
- Outcome specificity20/20Measures whether the reported benefit is concrete rather than a generic claim.
- Quantified impact25/252 metrics describe measurable impact.
- Deployment scale10/10Rewards evidence of a live or scaled deployment and meaningful integration scope.
Evidence confidence93%
- Source authority20/25The main evidence is an official vendor-published customer story.
- Claim traceability23/2511 of 12 key fields link to a named source.
- Direct evidence20/204 direct source quotes retained.
- Date provenance10/10The publication date comes from the source page or publisher.
- Source transparency10/101 source reference document publisher, URL, authority, retrieval, and supported fields.
- Implementation specificity10/10Rewards source-backed components, data flow, quantified impact, and a specific outcome.
Traceability
Sources & provenance
Field coverage shows which facts each source supports. Dates are shown only when documented by the source.
- https://www.capgemini.com/in-en/news/client-stories/aeta-shakes-up-earthquake-prediction-with-a-global-competition/CapgeminiFirst-party
AETA shakes up earthquake prediction with a global competition - Capgemini India
Supports- Customer
- Region
- Business problem
- Solution
- Products
- Solution components
- Data flow
- Integrations
- Business outcome
- Metrics
- Architecture
- evidence quotes
- Publication date
Date evidence
- Published on source page Aug 4, 2023 · Capgemini · First-partyOpen date source
Not separately sourced Industry, Company size. Treat these normalized attributes as lower-confidence until a dedicated source is added.
Related
Similar use cases
Capgemini Sample May 8, 2026 Generali GC&C Generali GC&C needed to simplify a complex underwriting and operations landscape spanning CRM, document management, and multiple country processes so underwriters could work faster with better data quality and lower… 1.9 TB of content migrated from multiple countries AGORA · Dynamics 365 · Gen AI-powered assistant 
Databricks Sample May 21, 2024 North Dakota University System (NDUS) For NDUS to compete in the fierce higher education market, each institution must work together; however, its data teams lacked the infrastructure to collaborate and scale its use cases. Lacking a modernized data and AI… 100% Reduced procurement time and costs leveraging Databricks on Azure Agent Bricks Capgemini Sample Dec 12, 2024 UNESCO UNESCO wanted to make large-scale learning-assessment data far easier for policymakers and educators to analyze so they could make better decisions about foundational reading outcomes. 251 million children around the world are out of school Amazon Bedrock · Amazon SageMaker · LangChain