Starky Interactive

Starky is an interactive bulldozer that represents the future of construction technology! With our automation features, voice activation, and safety features, we’re redefining construction!

Comment

Starky Interactive

Problem

Modern construction sites face three persistent challenges.

1. Labor Efficiency & Fatigue

Heavy machinery requires constant human operation. Operators must repeatedly adjust, reposition, and manually monitor grading tasks.

This leads to:

  • Operator fatigue
  • Slower work cycles
  • Inconsistent surface results
  • Extended project timelines

Repetitive manual corrections reduce productivity and increase physical strain.

2. Safety Blind Spots

Operators inside vehicles often have limited visibility around their equipment. Ground workers entering operational zones create collision risks.

This results in:

  • Increased accident potential
  • Delayed hazard awareness
  • Reactive instead of proactive safety systems

Many incidents occur due to lack of real-time proximity detection.

3. Communication Barriers

Construction sites are loud and chaotic. Verbal communication is difficult and signals can be misinterpreted.

This leads to:

  • Inefficient coordination
  • Delays in execution
  • Higher risk of operational mistakes

Despite automation advances in other industries, construction still relies heavily on manual, reactive control systems.

Inspiration

We were inspired by the gap between intelligent automation in industries like logistics and manufacturing and the limited sensing capabilities in traditional construction equipment.

We wanted to explore how a construction vehicle could:

  • Assist operators instead of relying entirely on them
  • Increase environmental awareness
  • Reduce safety risks
  • Improve efficiency through intelligent support systems

Starky is our prototype toward smarter, awareness-driven construction machinery.

What It Does

Starky Interactive is a smart mini-bulldozer that improves efficiency, safety, and machine awareness.

Automated Leveling

  • Dual bulldozing mechanisms push and smooth terrain
  • Reduces repeated manual corrections
  • Improves grading consistency

Real-Time Human Detection & Automatic Stopping

  • Uses OpenCV-based computer vision
  • Detects people in front of the vehicle
  • Automatically stops movement when a person is detected
  • Activates hazard LED lights

This helps reduce blind spot risks.

Efficient Movement

  • Optimized motor coordination
  • Distributes material evenly
  • Minimizes wasted passes

Starky transforms a basic vehicle kit into a semi-autonomous, safety-aware grading system.

How We Built It

We implemented a multi-component embedded system.

  • Computer running Python + OpenCV for computer vision and decision making
  • ESP32-CAM for live video capture
  • Arduino for motor and hardware control
  • Custom 3D-printed and laser-cut dual blade attachments
  • Integrated hazard LED safety system
  • Battery-powered onboard setup

System Flow

  1. ESP32-CAM streams video to a computer.
  2. The computer runs OpenCV to detect humans in the video feed.
  3. If a human is detected:
    • The computer sends a stop signal to the Arduino.
  4. The Arduino immediately:
    • Stops the motors
    • Activates hazard LEDs

Technologies Used

  • Python for computer vision and system logic
  • C++ for embedded motor control
  • OpenCV
  • Arduino platform

All mechanical and electrical components were designed and integrated by our team.

Challenges We Ran Into

  • Tuning blade geometry for consistent flattening
  • Managing motor torque under sand resistance
  • Synchronizing dual bulldozing mechanisms
  • Reducing camera latency
  • Handling serial communication timing
  • Power distribution across multiple boards
  • Ensuring fast and reliable emergency stop response

Integrating mechanical, electrical, and software systems simultaneously required continuous testing and refinement.

Accomplishments We're Proud Of

  • Integrated computer vision with real-time motor control
  • Implemented human detection with automatic stopping and hazard signaling
  • Designed and fabricated functional dual-leveling attachments
  • Achieved consistent surface smoothing performance
  • Built a safety-aware construction prototype within the deadline

We turned a basic robot chassis into a multi-system intelligent grading platform.

What We Learned

  • Construction automation requires both mechanical precision and intelligent software
  • Safety systems must be proactive and fast
  • Multi-device communication requires careful synchronization
  • Real-world testing exposes weaknesses quickly
  • Emergency stop systems must prioritize reliability over complexity

What's Next for Starky Interactive

  • Fully autonomous grading paths
  • Depth sensing for precision leveling
  • LiDAR-based terrain mapping
  • Adaptive blade adjustment based on resistance feedback
  • Advanced collision avoidance systems
  • Fleet coordination between multiple units
  • Real-time telemetry dashboards

Our long-term vision is construction equipment that actively assists operators, reduces labor strain, and improves safety through intelligent automation and environmental awareness.

Built With

Share this project:

Updates