In short
Semiconductor teams run several tools that each own one kind of data: a ticket tracker for tasks, source control for code, EDA tools for design results, PLM for product definition, ERP for parts and orders, and MES for the factory. None of them owns the program's date, which depends on all of them at once.
What does each tool own?
Each is the authority for its own data:
- Ticket trackers such as Jira: tasks, bugs and sprints.
- Source control such as Perforce or Git: the design and the code.
- EDA tools: regression, coverage and timing results.
- PLM: bills of materials, configurations and change orders.
- ERP: parts, purchase orders and deliveries.
- MES and the test floor: lots, yields and test data.
- Email and chat: what suppliers, the foundry and customers have committed to.
What falls between them?
Dependencies that cross tools. A change order in PLM that moves a build. A purchase order whose lead time has grown. A foundry email that moves a shuttle. Each is recorded correctly in its own tool and connected to nothing else.
Does a planning tool solve it?
A master schedule in a planning tool holds the plan. It is updated by people, so it is as current as the last person who updated it. It tells you what was intended, not what is true today.
What is the missing piece?
A layer that reads all of these, holds each commitment against its evidence, and acts when one moves. Nothing is replaced; every tool stays the source of truth for its own data. Tymeline holds the program to its date.