Chapter 6 · Practice 3 of 3
Keep a smaller plan B that still makes a complete project
Why it matters
For the single riskiest part of your project, write down a smaller version that would still give you a complete, defensible project if the ambitious version fails.
Examples:
- Ambitious: real-time Urdu speech commands. Plan B: a fixed set of voice commands, or text input with the same workflow.
- Ambitious: custom hardware sensor network. Plan B: one sensor plus simulated data for the rest.
- Ambitious: a new model that beats published results. Plan B: a careful comparison of existing models on local data.
Mention the plan B briefly in the proposal. It reassures the committee, and if you need it at the mid evaluation, switching is a planned decision rather than a failure.
Do
- Write a plan B for the riskiest component
- Agree with your supervisor when you would switch
Avoid
- Discovering in the last month that the core idea does not work