Autonomous Wound Treatment Robot
founder:	artem
depends on:	3d_printed_robotics_initiative
interested:
software license:
hardware license:

~~META: status = planning ~~

Late one evening, I arrived at an emergency department with a wound infection (7 days post-injury, clear signs: warmth, pain, redness). The waiting room was almost empty, two people total.

I sat by the door, exposed the wound, and asked the nurse: “Can you look at this for 5 seconds and tell me if it can wait until morning?”

She looked me in the eyes: “No. You're not from our district. Go to Prague 1.”

She could have lowered her head for 5 seconds. Instead, she sent me to reception for a taxi. It would have taken less time to assess the wound than to redirect me. The room was empty. Administrative boundaries mattered more than basic human compassion.

This project exists because:

• Emergency systems fail patients over bureaucracy
• Consistent, accessible first-line care shouldn't depend on luck
• Technology can provide the baseline care that humans sometimes refuse to give

An open-source 3D-printed robotic system for autonomous wound assessment and treatment.

Core Functions:

• AI vision assessment (infection detection, severity scoring)
• Ultrasonic wound debridement (bacteria elimination, biofilm removal)
• Automated cleaning and treatment application
• Treatment recommendations

Potential Applications:

• Emergency department triage (overnight/weekend shifts)
• Rural clinics with limited staff
• Veterinary clinics (easier regulatory pathway for initial testing)
• Field medicine (refugee camps, disaster zones)

Open Source Philosophy:

• All hardware designs (CAD files, STL files)
• All software (control systems, AI models, protocols)
• Full documentation for replication
• MIT/Apache license

Robotic Arm:

• 3D-printed structure (PETG/ABS for sterilization compatibility)
• 6-DOF design based on existing open-source arms
• Budget servo/stepper motors
• Tool changer mechanism for multiple attachments

Four Tool Attachments:

• Air nozzle — sterile compressed air (debris removal, drying)
• Liquid dispenser — saline/antiseptic spray system
• Ultrasonic probe — 20-40 kHz debridement head
• Ointment applicator — automated dosing system

Vision System:

• USB camera + depth sensor
• AI-based wound assessment
• Thermal imaging (optional, for infection detection)

Control:

• Raspberry Pi 4 or Arduino-based controller
• Force/distance sensors for safety
• Emergency stop mechanism

1. AI scan → assess wound (size, depth, contamination, infection signs)
2. Pre-cleaning → saline rinse + air debris removal
3. Ultrasonic debridement → antiseptic bath + ultrasound (30-60 sec)
   • Critical: ultrasound requires liquid medium to work
4. Final rinse → sterile saline flush
5. Air dry → compressed air
6. Apply ointment → levomekol or equivalent
7. Verification scan → check cleaning quality, repeat if needed
8. Output recommendation → antibiotic prescription yes/no

Vision & AI:

• OpenCV for image processing
• TensorFlow/PyTorch for ML models
• Wound segmentation (U-Net architecture)
• Infection classifier (CNN)

Control System:

• Custom kinematics or MoveIt integration
• Real-time force monitoring
• Safety collision detection
• Treatment protocol state machine

Data & Logging:

• All treatments logged for quality analysis
• Continuous model improvement from field data

• ☐ Design 3D-printable arm structure
• ☐ Source and test motors, sensors, components
• ☐ Build single-axis test rig
• ☐ Test each tool attachment independently:
  • ☐ Air nozzle pressure control
  • ☐ Liquid dispenser accuracy
  • ☐ Ultrasonic probe effectiveness
  • ☐ Ointment application consistency
• ☐ Build AI vision system (target: 70%+ accuracy on synthetic wounds)
• ☐ Create synthetic wound models for testing
• ☐ First complete treatment cycle demonstration

Milestone: Working prototype treats synthetic wounds with basic automation

• ☐ Assemble full 6-DOF robotic arm
• ☐ Implement tool changer mechanism
• ☐ Integrate all subsystems (vision, control, safety)
• ☐ Improve AI accuracy to 85%+ on diverse wound types
• ☐ Collect 100+ test treatments on synthetic models
• ☐ Document full build process for replication
• ☐ Explore partnerships:
  • ☐ Veterinary clinics (easier regulatory environment)
  • ☐ NGOs working in field medicine
  • ☐ University research collaboration

Milestone: System consistently treats wounds autonomously, documentation published

• ☐ Partner with veterinary clinic for supervised testing
• ☐ Collect real-world treatment data
• ☐ Refine protocols based on feedback
• ☐ Publish findings (blog posts, conference papers, videos)
• ☐ Build community of replicators
• ☐ Research regulatory pathways (veterinary first, then human)
• ☐ Explore grant opportunities for continued development

Milestone: 3+ external sites testing replicated systems, peer-reviewed validation

• ☐ Support multiple deployment sites
• ☐ Develop “enterprise” features for clinical integration
• ☐ Pursue medical device certification (if feasible)
• ☐ Expand to humanitarian applications
• ☐ Continue open-source development with growing community

Estimated component sources (not final):

Cost target: Keep total build under €5,000 for full system to enable widespread replication

Technical Goals:

• AI wound assessment accuracy >85%
• Treatment cycle time <10 minutes
• Zero safety incidents during testing
• System replicable by others following documentation

Community Goals:

• 2-3 active collaborators by Month 6
• Full documentation published by Month 12
• 3+ external replications by Month 24
• Published validation study (blog/paper/conference)

Impact Goals:

• Demonstrate viability of autonomous wound care
• Provide accessible healthcare option for underserved areas
• Inspire similar open-source medical robotics projects

Current Team:

• Project Lead: artem (systems integration, project coordination)

Looking For:

• Robotics engineer — arm design, kinematics, motor control
• ML/AI developer — vision system, wound classification models
• Medical advisor — protocol validation, safety review
• Embedded systems — microcontroller programming, sensor integration
• Documentation — technical writing, video tutorials, build guides
• Anyone interested! — part-time contribution welcome

How to Contribute:

• Join weekly robotics meetups at Brmlab
• Check GitHub repository (to be created)
• Join #robotics channel on Brmlab communication platform

Ultrasonic Wound Technology:

• SonicOne (Misonix) — clinical ultrasonic debridement
• UltraMIST — portable ultrasonic wound therapy
• QOUSTIC (Söring) — surgical ultrasonic systems

AI Wound Assessment:

• FDA-approved smartphone wound apps
• Academic research on diabetic ulcer classification
• Thermal imaging infection detection studies

Open-Source Robotic Arms:

• BCN3D Moveo — https://github.com/BCN3D/BCN3D-Moveo
• Thor — https://github.com/AngelLM/Thor
• SO-ARM100 — https://github.com/TheRobotStudio/SO-ARM100

Medical Robotics Inspiration:

• da Vinci Surgical System (tool changing mechanisms)
• STAR robot (autonomous suturing research)

Safety Measures:

• Force-limited actuators to prevent injury
• Patient-accessible emergency stop
• Human oversight for all treatments
• Automatic shutdown on error detection
• Sterile single-use tips for wound contact

Ethical Principles:

• Clear communication: system is an assistant, not a replacement for physicians
• Patient consent required before any treatment
• Edge cases automatically referred to human medical staff
• Privacy: minimal data collection, no storage without consent
• Accessibility: open-source ensures anyone can build and improve

Not Intended To:

• Replace physicians or trained medical professionals
• Handle complex medical cases
• Provide definitive medical diagnoses
• Operate without human oversight (initially)

Status: Planning phase — recruiting initial team

Immediate Next Steps:

1. [ ] Recruit 1-2 collaborators
2. [ ] Select base robotic arm design (BCN3D Moveo or Thor)
3. [ ] Source initial components (motors, camera, Raspberry Pi)
4. [ ] Create synthetic wound models for testing
5. [ ] Set up GitHub repository
6. [ ] Schedule first build session

First Meeting: TBD — announce on Brmlab calendar

Questions? Ideas? Want to help?

• Weekly meetings: Part of 3d_printed_robotics_initiative sessions
• Online discussion: Brmlab Slack/Discord #robotics
• GitHub: (repository link to be added)

Open Questions:

• Which robotic arm base should we use?
• Anyone have experience with ultrasonic systems?
• Contacts at veterinary clinics for future testing?
• Tips on medical device regulations in Czech Republic/EU?

• Parent project: 3d_printed_robotics_initiative
• GitHub repository: (to be created)
• Build documentation: (to be created)
• Contact: artem

Last Updated: 2026-01-18

License: Hardware designs and software will be released under MIT/Apache 2.0 open-source licenses