458  UAV Missions, Collision Avoidance, and Labs

458.1 Learning Objectives

By the end of this chapter, you will be able to:

  • Implement Collision Avoidance: Design Detect and Avoid (DAA) systems with appropriate separation standards
  • Select Mission Patterns: Choose appropriate trajectory patterns for different mission types
  • Plan Coverage Missions: Calculate flight parameters for area survey operations
  • Simulate UAV Networks: Build and test trajectory control algorithms in simulation environments

458.2 Prerequisites

Before diving into this chapter, you should be familiar with:

458.3 Collision Avoidance

458.3.1 Detect and Avoid (DAA) System

%% fig-alt: "Detect and Avoid (DAA) system flowchart: Sensors (cameras, radar, ADS-B, lidar) detect objects, tracking system predicts trajectories calculating closest point of approach, risk assessment evaluates time to collision and severity, three-level decision based on risk (low risk continues monitoring, medium risk alerts pilot, high risk executes avoidance), avoidance maneuvers include altitude change, lateral offset, speed adjustment, or hover/loiter, followed by mission resumption and continuous loop back to detection"
%%{init: {'theme': 'base', 'themeVariables': {'primaryColor': '#2C3E50', 'primaryTextColor': '#fff', 'primaryBorderColor': '#16A085', 'lineColor': '#E67E22', 'secondaryColor': '#ECF0F1', 'tertiaryColor': '#fff', 'nodeTextColor': '#2C3E50'}}}%%
graph TB
    subgraph "Detect and Avoid (DAA) System"
        Sensors["Sensors<br/>(Cameras, Radar,<br/>ADS-B, Lidar)"]

        Detect["Object Detection<br/>& Tracking"]

        Predict["Predict Trajectory<br/>(Closest Point<br/>of Approach)"]

        Assess["Collision Risk<br/>Assessment<br/>(Time to collision,<br/>Severity)"]

        Decision{Risk Level?}

        Monitor["Continue<br/>Monitoring<br/>(Low Risk)"]

        Alert["Alert Pilot<br/>(Medium Risk)"]

        Avoid["Execute Avoidance<br/>Maneuver<br/>(High Risk)"]

        Maneuver["Avoidance Actions:<br/>- Altitude change<br/>- Lateral offset<br/>- Speed adjustment<br/>- Hover/loiter"]

        Resume["Resume Mission"]
    end

    Sensors --> Detect
    Detect --> Predict
    Predict --> Assess
    Assess --> Decision

    Decision -->|Low| Monitor
    Decision -->|Medium| Alert
    Decision -->|High| Avoid

    Monitor -.->|Continuous| Detect
    Alert -.->|If escalates| Avoid

    Avoid --> Maneuver
    Maneuver --> Resume
    Resume -.->|Continue| Detect

    style Sensors fill:#2C3E50,stroke:#16A085,color:#fff
    style Assess fill:#E67E22,stroke:#2C3E50,color:#fff
    style Avoid fill:#E74C3C,stroke:#2C3E50,color:#fff
    style Resume fill:#16A085,stroke:#2C3E50,color:#fff

Figure 458.1: Detect and Avoid (DAA) system flowchart showing sensor input, trajectory prediction, risk assessment, and avoidance maneuver execution.
NoteWorked Example: Collision Avoidance Time Budget for Converging UAVs

Scenario: Two UAVs from different operators are flying in the same airspace. UAV-A detects UAV-B approaching on a collision course. You need to calculate the time available for detect-and-avoid (DAA) response and determine if the current separation standards are adequate.

Given: - UAV-A velocity: 15 m/s heading East - UAV-B velocity: 20 m/s heading North - Initial positions: UAV-A at (0, 0, 100m), UAV-B at (500m, -600m, 105m) - Minimum safe separation: 50 m - Sensor detection range: 400 m - DAA system latency: 2 seconds (detection + processing) - Avoidance maneuver: 5 m/s vertical climb

Steps: 1. Calculate closing geometry: UAV-A moves (+15, 0, 0) m/s, UAV-B moves (0, +20, 0) m/s. Relative velocity = UAV-B relative to UAV-A = (-15, +20, +0.167) m/s (including altitude convergence of 5m over ~30 seconds) 2. Calculate initial 3D separation: d = √(500² + 600² + 5²) = √(250000 + 360000 + 25) = 781.0 m 3. Calculate time to closest point of approach (CPA): - Position of UAV-A at time t: (15t, 0, 100) - Position of UAV-B at time t: (500, -600+20t, 105) - Distance² = (500-15t)² + (-600+20t)² + 5² - d(Distance²)/dt = 0 → -30(500-15t) + 40(-600+20t) = 0 - -15000 + 450t - 24000 + 800t = 0 → 1250t = 39000 → t = 31.2 seconds 4. Calculate CPA distance: At t=31.2s: - UAV-A: (468, 0, 100) - UAV-B: (500, 24, 105) - CPA distance = √(32² + 24² + 5²) = √(1024 + 576 + 25) = 40.3 m < 50 m safe threshold! 5. Calculate detection time: UAV-B enters 400m detection range when: √((500-15t)² + (-600+20t)² + 5²) = 400 - Solving: t ≈ 11.8 seconds (UAV-B detected at ~400m range) 6. Calculate available response time: Time from detection to CPA = 31.2 - 11.8 = 19.4 seconds. Minus 2 second DAA latency = 17.4 seconds for avoidance 7. Verify avoidance maneuver: UAV-A climbs at 5 m/s for 17.4 seconds = 87 m altitude gain. New altitude separation = 5 + 87 = 92 m (vertical) at CPA time. New 3D separation at CPA = √(32² + 24² + 92²) = 99.6 m > 50 m (safe!)

Result: Without avoidance, the two UAVs would pass within 40.3 m (collision risk). With 400 m detection range and 2-second DAA latency, the system has 17.4 seconds to execute avoidance. A 5 m/s vertical climb provides 92 m altitude separation, achieving 99.6 m total separation at CPA - safely above the 50 m minimum.

Key Insight: Collision avoidance is a time-budget problem. Detection range, processing latency, and maneuver capability must combine to provide sufficient separation before CPA. The critical formula is: Required detection range = Closing speed × (DAA latency + Maneuver time + Safety margin). For high-speed convergences (>30 m/s combined), 400 m detection range may be insufficient, requiring either longer-range sensors (radar vs camera) or reduced operating speeds in congested airspace.

458.3.2 Separation Standards

Scenario Minimum Separation Method
UAV-UAV (same swarm) 5-10 m Formation control
UAV-UAV (different operators) 50-100 m ADS-B, radar
UAV-Manned Aircraft 500 m+ ADS-B, TCAS
UAV-Obstacle (building) 10-30 m Lidar, camera
UAV-Ground 30-120 m Altitude hold

458.4 Mission Types and Trajectory Patterns

458.4.1 Common Mission Profiles

Mission Pattern Key Considerations
Area Survey Lawnmower/Boustrophedon Overlap for complete coverage
Perimeter Patrol Circular/Racetrack Continuous monitoring
Point Inspection Hover + Spiral High-resolution data
Search & Rescue Expanding square Prioritize likely areas
Delivery Point-to-point Obstacle avoidance
Relay/Comm Station-keeping Minimize position error

458.4.2 Lawnmower Pattern for Area Coverage

%% fig-alt: "Lawnmower survey pattern showing UAV starting at corner, flying parallel tracks across survey area (Track 1 west to east, 180-degree turn, Track 2 east to west, turn, Track 3 west to east, turn, Track 4 east to west), with parameters of 80m altitude, 10 m/s speed, 115m track spacing with 20% overlap, completing survey and returning to base"
%%{init: {'theme': 'base', 'themeVariables': {'primaryColor': '#2C3E50', 'primaryTextColor': '#fff', 'primaryBorderColor': '#16A085', 'lineColor': '#E67E22', 'secondaryColor': '#ECF0F1', 'tertiaryColor': '#fff', 'nodeTextColor': '#2C3E50'}}}%%
graph TB
    subgraph "Lawnmower Survey Pattern"
        Start["Mission Start<br/>(Corner)"]

        Track1["Track 1<br/>(West to East)"]
        Turn1["180° Turn"]
        Track2["Track 2<br/>(East to West)"]
        Turn2["180° Turn"]
        Track3["Track 3<br/>(West to East)"]
        Turn3["180° Turn"]
        Track4["Track 4<br/>(East to West)"]

        Complete["Survey Complete"]
        Return["Return to Base"]

        Params["Parameters:<br/>- Altitude: 80m<br/>- Speed: 10 m/s<br/>- Track spacing: 115m<br/>- Overlap: 20%"]
    end

    Start --> Track1
    Track1 --> Turn1
    Turn1 --> Track2
    Track2 --> Turn2
    Turn2 --> Track3
    Track3 --> Turn3
    Turn3 --> Track4
    Track4 --> Complete
    Complete --> Return

    Params -.->|Configure| Track1

    style Start fill:#2C3E50,stroke:#16A085,color:#fff
    style Track1 fill:#E67E22,stroke:#2C3E50,color:#fff
    style Track2 fill:#E67E22,stroke:#2C3E50,color:#fff
    style Complete fill:#16A085,stroke:#2C3E50,color:#fff

Figure 458.2: Lawnmower survey pattern showing parallel tracks with 180-degree turns for systematic area coverage.

458.5 Hands-On Lab: UAV Mission Planning

458.5.1 Lab Activity: Coverage Mission Design

Scenario: Plan a UAV survey mission for a 1 km × 1 km agricultural field.

Given: - UAV: DJI Mavic-class (30 min flight time, 15 m/s max speed) - Camera: 84° FOV, need 5 cm/pixel resolution - Wind: 5 m/s from the north

Tasks:

  1. Calculate Flight Altitude
    • For 5 cm/pixel with given sensor, altitude ≈ 80 m
  2. Calculate Track Spacing
    • At 80 m altitude, swath width = 2 × 80 × tan(42°) ≈ 144 m
    • With 20% overlap: track spacing = 144 × 0.8 = 115 m
  3. Calculate Number of Tracks
    • Field width: 1000 m ÷ 115 m = 9 tracks
  4. Calculate Flight Distance
    • 9 tracks × 1000 m + 8 turns × 50 m = 9,400 m
  5. Calculate Flight Time
    • At 10 m/s: 9,400 ÷ 10 = 940 s ≈ 16 minutes
    • Add takeoff/landing: ~20 minutes total
  6. Verify Feasibility
    • 20 min < 30 min battery → Mission feasible with 33% reserve
Parameter Value Calculation
Flight altitude 80 m Resolution requirement
Track spacing 115 m 144 m × 0.8 overlap
Number of tracks 9 1000 m ÷ 115 m
Total distance 9.4 km (9 × 1000) + (8 × 50)
Flight time ~16 min 9400 m ÷ 10 m/s
Battery reserve 47% 14 min remaining

Result: Mission is feasible with good safety margin.

458.6 Knowledge Check

Test your understanding of these architectural concepts.

Question 1: A UAV swarm of 10 UAVs monitors a wildfire. Fire spreads north; UAVs must reposition. What coordination approach is most appropriate?

💡 Explanation: UAV swarm coordination strategies trade off centralization vs autonomy: Centralized control (best choice): Ground Control Station (GCS) has: (1) Global view: Sees all UAV positions, fire boundary, no-fly zones. (2) Computational power: Runs optimization algorithms (assign UAVs to maximize fire perimeter coverage). (3) Human oversight: Safety-critical missions require human approval. (4) Communication with authorities: Coordinates with firefighters, air traffic control. Algorithm: Receive fire map from UAVs → Calculate optimal coverage positions → Send waypoint commands to each UAV → UAVs navigate to assigned positions. Ensures coordinated repositioning (no collisions, optimal coverage). Why not alternatives? Independent local decisions: Each UAV sees only local sensor data. Results in: clustering (all UAVs follow fire, leaving gaps), conflicts (multiple UAVs target same area), suboptimal coverage. Example: 3 UAVs detect fire north, all move north leaving south uncovered. Democratic voting: Slow (requires multi-round consensus), doesn’t guarantee optimal solution, communication overhead (10 UAVs = 45 pairwise links for voting). Manual repositioning: Defeats automation purpose, too slow (fire spreads faster than manual redeployment). Hybrid approach (best for large swarms): Hierarchical - GCS assigns regions to UAV clusters, cluster leader coordinates local UAVs autonomously. Balances global optimization with local autonomy, scales to 100+ UAVs, resilient to GCS communication loss (local clusters continue operating).

Question 2: A UAV mapping mission uses a camera footprint of 144 m and requires 20% side overlap between adjacent flight tracks. Approximately what track spacing should you plan?

💡 Explanation: B. With 20% overlap, spacing = footprint × (1 − overlap) = 144 m × 0.8 ≈ 115 m. This ensures coverage without leaving gaps while avoiding unnecessary extra flight distance.

Question 3: You have many candidate waypoints for data collection. Which approach is commonly used to quickly reduce total route length without solving the full optimal TSP exactly?

💡 Explanation: C. Exact TSP solutions can be expensive for large waypoint sets. Practical mission planners often use heuristics (nearest-neighbor, 2-opt) to produce good-enough routes that reduce distance and energy consumption.

Question 4: In multi-UAV formation control, what is the primary purpose of consensus-based coordination algorithms?

💡 Explanation: A. Consensus methods let distributed agents converge on common values using local neighbor communication (e.g., velocity alignment, spacing). This supports stable formations, coordinated motion, and collision avoidance even when the swarm is decentralized.

458.8 Summary

This chapter covered collision avoidance, mission types, and hands-on UAV mission planning:

  • Detect and Avoid (DAA): Multi-sensor systems (cameras, radar, ADS-B, lidar) detect objects, predict trajectories, assess collision risk, and execute avoidance maneuvers when needed
  • Separation Standards: Different scenarios require different minimum separations - from 5-10 m for same-swarm UAVs to 500+ m for UAV-manned aircraft encounters
  • Mission Patterns: Different missions use different trajectory patterns - lawnmower for area survey, circular for perimeter patrol, expanding square for search and rescue
  • Coverage Planning: Flight altitude, track spacing, and overlap requirements determine mission feasibility within battery constraints

Deep Dives: - UAV Fundamentals and Topologies - Network architectures - UAV Production and Review - Complete system implementation - UAV FANET Integration - Protocol integration

Comparisons: - Mobile Ad-Hoc Networks - Ground vs aerial mobility - Wireless Sensor Networks - Static vs mobile collection

Applications: - Search and Rescue - Emergency missions - Precision Agriculture - Crop monitoring

Design: - Energy-Aware Design - Battery optimization - Network Simulation - Testing tools

Learning: - Simulations Hub - UAV flight simulators - Videos Hub - Mission planning tutorials

The following AI-generated figures provide alternative visual representations of concepts covered in this chapter. These “phantom figures” offer different artistic interpretations to help reinforce understanding.

458.8.1 Additional Figures

U T M System diagram showing key concepts and architectural components

U T M System

458.9 What’s Next

Having explored UAV trajectory control, energy-aware mission planning, swarm coordination, and collision avoidance, the next chapter examines ad hoc network routing protocols and simulation techniques for mobile IoT systems.

Continue to Ad Hoc Networks: Labs and Quiz →