Factory Maintenance Role-Playing Swarm using CrewAI
Complex manufacturing plants require seamless coordination between distinct human operational personas—such as Plant Maintenance Managers, Reliability Engineers, and Procurement Leads. In this challenge, developers will utilize CrewAI to orchestrate an autonomous team of role-playing AI agents that collaborate to evaluate predictive maintenance reports and negotiate equipment downtime windows. The CrewAI swarm consists of a Maintenance Diagnostic Agent, a Production Scheduler Agent, and a Parts Procurement Agent. Given predictive failure signals indicating impending motor breakdown, the agents engage in role-play deliberation to schedule repairs without disrupting scheduled factory output target deadlines.
What you are building
The core problem, expected build, and operating context for this challenge.
Orchestrate a role-playing agent swarm using CrewAI to negotiate factory downtime windows and coordinate preventive maintenance.
How work is evaluated
Evaluates the consensus scheduled window generated by the CrewAI agents against plant constraints.
Challenge datasets
Public datasets, plus the private datasets you have access to.
How submissions are scored
These dimensions define what the evaluator checks and which criteria separate a passable run from a strong one.
Optimal Shift Selection
Maintenance must be scheduled during the lower demand shift (Shift 2)
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
Agent Consensus Efficiency
Number of turns required for crew consensus • target: 3 • range: 1-10
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
What you should walk away with
Define CrewAI agents with specific roles, goals, and operational backstories
Orchestrate sequential and hierarchical tasks across the agent crew
Implement inter-agent negotiation tools for downtime window optimization
Produce a unified preventive maintenance strategy document
Reference links and supporting material
Synthetic shift production demand tables and equipment failure probability metrics for press shop manufacturing operations.
Challenge setup
Evaluates the consensus scheduled window generated by the CrewAI agents against plant constraints.
Challenge input
JSON with failure_risk_pct, target_machine, production_shift_demand
CrewAI
Assigned primary agent framework
Evaluated output
JSON with scheduled_shift, downtime_hours, approved_by_crew (boolean)
- Maintenance must be scheduled during the lower demand shift (Shift 2)
- Number of turns required for crew consensus • target: 3 • range: 1-10
- Benchmark: Multi-Agent Consensus Scheduling Benchmark
- Agent Consensus Efficiency target: 3
- 1 public reference case
- Python execution harness
- Python sandbox (unavailable on Versalist)
Summary shows the first task, up to three tools, and up to three evidence items. Open the technical recipe to view all configured items.
View technical recipe
Configured tools
- CrewAI · Required
- crewAI · Optional
Evaluation contract
- Optimal Shift Selection · Weight 1
- Agent Consensus Efficiency · Weight 1
Metrics
- Agent Consensus Efficiency · Target 3
Benchmarks
- Multi-Agent Consensus Scheduling Benchmark
- general
Reference cases
- 1 public reference case
Workflow tasks
- Task 1
Input: JSON with failure_risk_pct, target_machine, production_shift_demand
Output: JSON with scheduled_shift, downtime_hours, approved_by_crew (boolean)
Execution harness: Python
Evaluation tools
- CrewAI · Role-playing multi-agent orchestrator
Recipe freeze marker
No recipe freeze time is recorded.