Design a Robot System
You have studied the five subsystems of a robot, the physics of sensing and actuation, the demands of real-time computation, the discipline of reliability engineering, and the art of systems tradeoffs. Now you put it all together. This lesson is a complete engineering design exercise: you will select a mission, derive requirements, design every major subsystem, define interfaces, identify tradeoffs, and assess failure modes. The goal is not a perfect robot — it is a defensible, coherent design where every choice is justified by analysis rather than assumption.
The Design Process
Professional robot design follows a structured process that applies whether you are designing a hobbyist robot or an industrial system. The steps are: 1. Mission definition: What must the robot do? Where? Under what conditions? For how long? For whom? The mission statement is written in terms of observable outcomes, not internal mechanisms. 2. Requirements derivation: From the mission, derive specific, measurable, verifiable requirements. 'The robot must navigate reliably' is not a requirement. 'The robot shall navigate a 100 m corridor within 2 minutes with fewer than 0.1% collision events per mission' is a requirement. 3. Concept of operations (ConOps): Describe in detail how the robot will be operated — how it is deployed, monitored, recharged, and maintained. Many requirements that are missed in early design emerge from a careful ConOps. 4. Subsystem design: For each of the five subsystems (sensing, actuation, computation, power, structure), select specific components and justify the selection against the requirements. 5. Interface definition: Specify every interface between subsystems. What signals pass? At what rate? In what format? Under what error conditions? 6. Tradeoff documentation: Identify the three to five most significant design tradeoffs, state the alternatives considered, and justify the chosen option. 7. Risk assessment: Apply FMEA to the three highest-risk failure modes and specify the fault detection and recovery strategy for each. This is not merely a planning exercise — at each stage you may discover that earlier decisions must be revised. Design is inherently iterative.
The most common design error made by beginners is choosing components before writing requirements — picking a cool motor or a shiny sensor, then trying to justify it. Professional engineers write requirements first. Requirements tell you what success looks like; they are the measuring stick against which every design decision is evaluated.
Mission Options
Choose one of the following missions for your robot design. Each has different dominant constraints that will drive your subsystem choices. Mission A: Hospital Medication Delivery Robot. Operates indoors in a hospital environment. Must navigate corridors, operate elevators, and deliver medication carts to nursing stations autonomously. Must operate safely near patients and staff, including children and people in wheelchairs. Must operate 20 hours per day, 7 days a week. Mission B: Coastal Water Quality Monitoring Drone. An unmanned surface vehicle (USV) that autonomously navigates coastal waters, collecting water samples and sensor measurements (pH, turbidity, temperature, dissolved oxygen) at designated waypoints. Must survive 1 m wave heights and operate up to 6 hours per deployment. Mission C: Fruit Harvesting Robot. Operates in an orchard, detecting and harvesting ripe fruit from trees without bruising it. Must work in variable outdoor lighting, withstand light rain, and operate 12 hours per day during harvest season. Mission D: Search and Rescue Scout. A small legged robot that enters a collapsed building to map passable pathways and locate survivors using thermal imaging and audio, then relays the map to human rescue teams. Must fit through a 0.4 m x 0.4 m opening.
Full Robot System Design
- Complete all seven stages of the robot design process for your chosen mission. This is your capstone project for the module.
- STAGE 1 — MISSION DEFINITION (1 paragraph)
- State your chosen mission in precise terms. Include: what the robot must do, the operating environment, the duty cycle (how long per day), any human proximity requirements, and the target service life.
- STAGE 2 — REQUIREMENTS (minimum 8 requirements)
- Write at least 8 specific, measurable, verifiable system requirements. Each requirement must state: the subject (the robot shall...), the condition (when operating in...), and the measurable criterion (within X, at least Y, no more than Z). Include at least one requirement for each subsystem.
- STAGE 3 — CONCEPT OF OPERATIONS (1-2 paragraphs)
- Describe how the robot is deployed, monitored during operation, recovered when something goes wrong, recharged or refueled, and maintained. Identify who the operators are and what their skill level is assumed to be.
- STAGE 4 — SUBSYSTEM DESIGN (one paragraph per subsystem)
- For each of the five subsystems, name specific components you would use (real or representative), state the key performance parameters, and explain why these choices satisfy your requirements. Address all five: sensing, actuation, computation, power, structure.
- STAGE 5 — INTERFACE DEFINITION (interface table)
- Create a table with one row per subsystem interface. Columns: Interface Name, Subsystem A, Subsystem B, Signal/Energy Type, Rate or Capacity, Format or Protocol, Key Constraint or Risk.
- STAGE 6 — TRADEOFF ANALYSIS (three tradeoffs)
- Identify the three most significant design tradeoffs in your design. For each: name the tradeoff, describe the alternatives you considered, explain the quantitative or qualitative factors that drove your choice, and state what you gave up.
- STAGE 7 — RISK ASSESSMENT (three failure modes)
- Apply FMEA to your three highest-severity failure modes. For each: describe the failure mode, its effect on the robot and on its human environment, its severity (1-10), its occurrence likelihood (1-10), its detectability (1-10), the computed RPN, your proposed detection mechanism, and your proposed recovery action.
- Present your complete design to the class. Be prepared to defend every choice against alternative proposals.
An excellent design is not one where everything is perfect — it is one where every constraint is understood, every tradeoff is documented, and every risk has a mitigation. A design that says 'we chose aluminum over carbon fiber because CFRP costs 10x more and our mission tolerates 0.8 kg extra structure mass within our 15 kg budget' is far more valuable than one that chose CFRP because 'it is the best material.' Show your reasoning.
A student writes the following system requirement: 'The robot should navigate well in the hospital.' A systems engineer rejects it. What is wrong with this requirement and what would a correctable version look like?