Skip to main content
Loading navigation
Back to challenges

Google ADK Multimodal Casino Floor Traffic & Revenue Optimization

Build a Google ADK multimodal agent using Gemini to analyze floor plan imagery and IoT sensor streams for regional casino resorts. The system identifies high-conversion non-gaming revenue layouts and predicts spatial yield optimization with 88% precision.

Business OperationsHosted by Vera
Sign in free to start

What you are building

Uses Google Agent Development Kit (ADK) and Gemini multimodal capabilities to process floor layout diagrams and real-time foot-traffic heatmaps for non-gaming resort revenue optimization.

How work is evaluated

Evaluation

Evaluates multimodal layout parsing and non-gaming revenue prediction accuracy.

Datasets

Public datasets, plus the private datasets you have access to.

How submissions are scored

Dimensions
2 scoring checks
Binary
2 pass or fail dimensions
Ordinal
0 scaled dimensions
Dimension 1

Revenue Metric Bounds

Verifies that projected revenue lift values are physically realistic (0-50%)

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Dimension 2

Layout Recommendation Relevance

Expert rating score on space reallocation suggestions • target: 0.88 • range: 0-1

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Learning goals

  • Implement Google ADK multimodal processing agents leveraging Gemini APIs

  • Process visual floor plans alongside tabular spatial foot-traffic sensors

  • Identify optimal dining, retail, and entertainment layout placements

  • Predict non-gaming square-foot yield optimizations

Resources

Dataset notes

20 regional casino floor plans with paired foot-traffic heatmaps and venue yield metrics.

Challenge setup

Evaluates multimodal layout parsing and non-gaming revenue prediction accuracy.

Challenge input

JSON containing layout_image_uri, hourly_foot_traffic_count, and current_non_gaming_sqft

Google ADK

Provides direct multimodal integration with Gemini and Vertex AI.

Evaluated output

JSON containing recommended_reallocation, projected_revenue_lift_pct, and bottleneck_zones

Evaluation checks
  • Verifies that projected revenue lift values are physically realistic (0-50%)
  • Expert rating score on space reallocation suggestions • target: 0.88 • range: 0-1
Configured evaluation evidence
  • Benchmark: MultimodalSpatialYield-v1
  • Layout Recommendation Relevance target: 0.88
  • 1 public reference case
Configured runtime
  • Python execution harness
  • Python sandbox (unavailable on Versalist)

Shows the first task, up to three tools, and up to three evidence items.

View technical recipe

Configured tools

Environment
  • Modal · Optional
Action Space
  • Google ADK · Required
  • Google · Optional
Policy Serving
  • Modal · Optional

Evaluation contract

  • Revenue Metric Bounds · Weight 1
  • Layout Recommendation Relevance · Weight 1

Metrics

  • Layout Recommendation Relevance · Target 0.88

Benchmarks

  • MultimodalSpatialYield-v1
  • general

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON containing layout_image_uri, hourly_foot_traffic_count, and current_non_gaming_sqft

    Output: JSON containing recommended_reallocation, projected_revenue_lift_pct, and bottleneck_zones

    Execution harness: Python

Evaluation tools

  • Google ADK · Agent Development Framework

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Google ADK Multimodal Casino Floor Traffic & Revenue Optimization