Problem 8
Catch the shortage before the line stops, then substitute, source, or resequence so the day still runs.
The Production Planning module of a manufacturing ERP schedules production runs based on sales orders and raw-material availability. When a raw material falls short, a planner has to manually decide whether to delay the run, substitute a material, or place an urgent purchase order — usually under time pressure, right before a line would otherwise stop.
Manual replanning is slow and reactive, and it rarely explores substitute materials or multiple supplier options before committing to an urgent, often expensive, purchase. A shortage on one material can also unnecessarily stall other runs that could have been resequenced ahead of it.
The challenge is to build an agent that catches shortages early and resolves them with the least disruption to the overall schedule.
This problem is relevant to organizations that plan and run physical production, including:
Build an agentic AI module that checks a production schedule against raw-material stock in a mock ERP, detects shortages that will block a scheduled run, resolves them by proposing an approved substitute material or by sourcing and negotiating an urgent purchase with a supplier agent, and re-sequences the schedule to minimize delay.
Participants are not required to model every planning rule. Teams may choose the sourcing and rescheduling scope they want to address and are encouraged to innovate beyond the examples provided.
Shortage Detection
Cross-check the production schedule's material requirements against current stock and open purchase orders.
Substitution Reasoning
Identify approved substitute materials that satisfy the same production specification.
Urgent Sourcing & Negotiation
Source and negotiate price and lead time with one or more mock supplier agents when no substitute is viable.
Schedule Re-Sequencing
Reorder unaffected production runs ahead of a blocked one to minimize total idle time.
Level 1 (Basic Agency)
The agent reads a CSV of the day's production schedule and raw-material stock, and flags which scheduled runs will be blocked by a shortage and by how much, with reasoning.
Level 2 (Tool Integration)
For each shortage, the agent checks a substitute-materials table and recommends a substitute when compatible; otherwise it auto-drafts an urgent purchase request to a mock "Supplier API." Level 3 (Multi-Agent Negotiation — Stretch Goal) A "Planner Agent" negotiates lead time and price with 2–3 "Supplier Agents" for urgent shortages that have no substitute, then re-sequences the full day's schedule to minimize total production delay and writes the updated schedule back into the mock ERP.
A working demo showing the shortage detection, resolution decisions, and updated schedule.
A generated result may include:
Participants are free to decide:
Teams may use synthetic or publicly available manufacturing datasets where appropriate.
1. Meaningful AI Usage
The solution should demonstrate why intelligent sourcing and rescheduling are required. Solutions should go beyond a single fixed "delay the run" default.
2. Detection and Resolution Should Be Connected
The specific shortage and its severity should meaningfully determine whether a substitute, an urgent order, or a resequencing is chosen.
3. Delay Minimization
The solution should attempt to minimize total production delay across the schedule, not just resolve the first shortage found.
4. Appropriate Handling of Different Cases
Teams should show the agent behaving sensibly when a substitute exists versus when an urgent order is the only option.
5. Actionable Output
The updated schedule should be structured so it could be handed directly to the shop floor.
6. Efficiency
Participants should be mindful of processing time and negotiation rounds given the short hackathon window — a well-scoped Level 1 or 2 is better than an unfinished Level 3.
7. Innovation Freedom
Participants are not restricted to a specific model, framework, scheduling algorithm, or technology.
What It Measures
Correctness of identified shortages and their severity
Soundness of substitution or sourcing decisions
How effectively the re-sequenced schedule reduces total delay
Ability to handle multiple simultaneous shortages
Whether AI provides meaningful value beyond a fixed delay rule
Architecture, implementation, and robustness
Sensible use of models and negotiation rounds within the time limit
Novelty of the sourcing or rescheduling approach
Ability to clearly show detection, resolution, and final schedule
Ability to evolve into a practical Production Planning add-on
Participants are free to choose their own technical approach.
Solutions may use any relevant combination of:
The listed technologies are examples only and should not limit innovation.
Given a production schedule and raw-material constraints, intelligently detect shortages, resolve them with the least disruption, and keep the overall schedule as efficient as possible.