Response principles.
Before any specific procedure: the framework that makes specific procedures work. Emergencies are stressful, and stress narrows attention. Cohort default response patterns are designed to keep operators in working mental state — not removing the stress, but making sure the response happens despite it. Four principles run through every section that follows.
The four response principles:
The "this won't happen to me" trap
A later cohort surveyed graduates 18 months post-graduation. Ten of fifteen graduates reported at least one situation they considered a real emergency during their first year. The categories distributed roughly evenly: in-flight failures, crashes in difficult terrain, weather emergencies, equipment issues. One operator in five experienced a bystander-related incident requiring careful response.
The data is clear: emergencies happen. Operators who treat them as inevitabilities — and prepare accordingly — handle them better than operators who treat them as remote possibilities. Cohort default training includes regular emergency drills throughout the program, not as a one-time checkbox.
If you're reading this page thinking "I'm too careful for this to apply to me" — that's the most dangerous mental state. The careful operators have emergencies too; they just handle them better because they've prepared.
What "calm" actually means in this context. Cohort instructors specifically distinguish between "feeling calm" (emotional state, not always available) and "acting calm" (behavioural pattern, always available with practice). You may feel terrified during an emergency; that's human. Acting calm — speaking deliberately, moving methodically, following the procedure — is a learned skill that works regardless of internal feeling.
Practical markers of acting calm:
- Speaking in complete sentences, not fragments.
- Standing still or moving deliberately, not pacing or running.
- Following the procedure step-by-step, not skipping ahead.
- Asking for help when needed; accepting help when offered.
These behaviours can be performed even when feeling stressed. Emergency drill repetition makes them the default response; without practice, the default tends toward fragmented action. This is why cohort training drills emergencies before graduates need them.
In-flight failures.
The drone is in the air. Something just changed — link quality dropped, GPS sat count dropped, a motor sounds wrong, the OSD shows an unexpected mode. You have seconds to decide whether this is "monitor closely" or "abort now". This section covers the specific failures and the response pattern for each.
The general response pattern for in-flight failures:
- Identify what changed. Voltage? GPS? Link? Mode? Visual contact? Name it before deciding.
- Take manual control if able. Switch out of mission mode to Position Hold or Angle. The drone now responds to sticks, not to the autonomous logic that may have caused the problem.
- Bring it back. Either fly it home manually, or trigger RTH if RTH is reliable for the current conditions.
- Land safely. Original launch zone or pre-identified emergency landing zone.
- Diagnose on the ground. Don't troubleshoot in the air; that's how mistakes compound.
The "fly-away" scenario
The most concerning in-flight emergency: drone heading off in an uncontrolled direction, not responding to RC input, not returning home. Possible causes: GPS confusion + autonomous mode logic failure, severe RC interference, FC firmware bug, magnetic interference confusing the compass.
Cohort default response sequence:
- Try Angle mode — disables GPS dependence; the drone should respond to sticks immediately.
- If no response, try the throttle-cut switch — configured in fc-setup.html as the kill switch. The drone falls; this is acceptable if it's heading toward a populated area or restricted airspace where the alternative is worse.
- If neither response, the drone is genuinely fly-away. RC failsafe (configured for RTH) will eventually trigger when the drone exits radio range. Track its likely path; prepare to recover or report.
Fly-aways are rare but not theoretical — a later cohort saw one (firmware bug, recovered intact). The kill-switch decision is genuinely hard: dropping a drone is bad, but a drone heading toward a populated area is worse. The decision should be made before the situation arises; verbalising "I will use the kill switch if X" during pre-flight (Section 3 of field-ops.html) commits to the decision in advance.
The diagnostic-on-the-ground principle. When something goes wrong in flight, the temptation is to troubleshoot from the ground station — try a different mode, push different sticks, see if it recovers. This is how minor problems become major ones.
The cohort default: get the drone on the ground first, then diagnose. Even if "trying one more thing" might recover the situation, the cost of failure (more serious crash, longer fly-away, potential injury) is much higher than the cost of a precautionary landing. Never sacrifice a known-safe landing opportunity for a chance to recover the mission.
Crash response.
Despite all precautions, sometimes the drone goes down. The crash response procedure is about three things: making sure no one is hurt, recovering the drone safely, and documenting what happened. Different scenarios (water, dense canopy, hard ground, structures) need different recovery approaches; injury risk varies similarly.
The crash response sequence (regardless of crash type):
- Verify no one is hurt. Visual scan of the crash area; ask cooperative members if anyone was nearby. If injury suspected, see Section 4 (bystander emergencies).
- Disconnect power if accessible. Even a crashed drone can have ESC/motor activity if the FC is still powered. A still-powered LiPo in a crash is also a fire risk (Section 5).
- Document the scene. Photos of the crash site before moving anything. The wreckage location, orientation, surrounding terrain. Useful for post-incident analysis and for cooperative records.
- Recover when safe. Different terrain types have different recovery procedures (below).
- Notify cooperative leadership. Honest brief explanation of what happened. Don't hide it; cooperatives respect honesty about problems.
The "should I keep flying today" decision
After a crash, the question of whether to continue the day's missions arises. Default cohort answer: no. Reasons:
- Equipment confidence: even with a backup drone, the cause of the crash is often unclear immediately. Resuming with potentially-related conditions risks a second crash.
- Operator state: a recent crash affects judgment. Even calm operators take time to reset; flying with degraded judgment is how second crashes happen.
- Cooperative relationship: the cooperative just witnessed something going wrong. Resuming immediately can feel reckless to them; rescheduling demonstrates care for the relationship.
Exceptions exist: minor crashes with clear causes (prop strike from a bird, hard landing in cohort default conditions) and confident operators sometimes continue. But the default — stop flying for the day after any crash — is the cohort recommendation. The mission gets rescheduled; the relationship and equipment stay intact.
If the crash was during a paid mission: the cooperative is paying for survey work that just got disrupted. Cohort default approach to communicate this:
- Honest acknowledgment: "I had a crash; the equipment is damaged; I won't be able to complete today's mission."
- Specific make-good: "I'll return with replacement equipment within X days at no additional cost. The original mission scope and pricing will apply."
- Documentation: provide a brief incident note within 24 hours so the cooperative has it for their records.
- No defensiveness: don't blame conditions, equipment, or anything external. The cooperative invited you; the equipment is your responsibility.
Later cohorts have examples of graduates whose handling of a crashed-mission situation strengthened the cooperative relationship: the cooperative observed honesty and professionalism under pressure, which is rare and valuable. The crash itself was the smaller event compared to the recovery handling.
People and bystanders.
Most cohort emergencies are equipment-only — no one is hurt; the worst case is replacing a drone. Some emergencies involve people. This section is the most important; getting it right matters more than any other emergency category. The procedures below cover injury response, hostile encounters, and animal incidents.
Important context
The procedures below are for cohort default operations and do not replace formal first aid training, professional medical guidance, or emergency services. Cohort program guidelines require operators to maintain basic first aid certification (PRC-recognised programs); this page builds on that foundation. For any serious injury, contacting local emergency services is the only correct first step. The Philippine Red Cross national hotline is 143; barangay-level emergency contact should be confirmed during pre-mission setup (Section 2 of field-ops.html).
Bystander injury from drone:
- Stop all flying immediately. Land any drone in the air; secure all equipment. The mission is over.
- Get to the injured person quickly but calmly. Running can cause additional injury or panic; walking with purpose is the right pace.
- Assess the injury. Use whatever first aid training you have. Don't move the person if neck/back injury is possible; let medical professionals make that call.
- Call for help. Cooperative manager often has barangay-level contacts. The barangay captain or local health worker may be the closest medical resource. Philippine Red Cross 143 for serious situations.
- Notify cooperative leadership immediately. They need to know to support response and inform local authorities if appropriate.
- Stay with the injured person until help arrives. Don't leave them alone; ensure transport to medical care is happening.
- Document afterwards, but not before care is in hand. The incident report comes after the person is safe and being treated.
Hostile encounter or refused access
Occasionally, someone (not part of the cooperative) confronts the operator about drone activity — concerned about privacy, surveillance, or perceived intrusion. This is rare in cohort agricultural contexts but worth being prepared for.
Cohort default response:
- Stop flying. Land the drone immediately; don't continue while the situation is unresolved.
- Speak respectfully. Acknowledge the concern; explain the work briefly and honestly. "We're doing a survey for the cooperative manager [name]; we're not photographing private property; happy to show you what we're capturing."
- Get cooperative leadership involved if the situation persists. The barangay captain or cooperative manager has standing the operator does not.
- Be prepared to leave if the situation can't be resolved. The mission can be rescheduled; an escalating confrontation cannot be undone.
- Don't engage aggressively, ever. Even if the other party is wrong about the work being intrusive, escalation serves no one.
Documentation matters here too. Write down what happened, who was involved, how it resolved. This becomes part of the incident record (Section 6).
Animal incidents are a less-discussed category but happen. Most often: aggressive dog approaching the operator or a working drone area. Less common: livestock spooked into stampede; raptor attacking the drone in flight (carabao egret, sea eagle).
Practical responses:
- Aggressive dog: don't run. Stand still, avoid eye contact, speak calmly. Cooperative members usually know the dog and can call it back. Most farm dogs are protecting their territory rather than attacking; backing away slowly resolves it.
- Spooked livestock: prioritise getting bystanders to safety. Don't try to control livestock; cooperative members know how. Land the drone away from the stampede path.
- Bird attack on drone: rare but documented. Land if possible; the bird is defending territory. Continuing to fly while attacked usually causes a crash (and may injure the bird).
A later cohort had one bird-attack incident (sea eagle in coastal cooperative) that ended in a controlled landing. No injury to bird or drone. The lesson: birds defend their territory from drones; surveys near eagle territory benefit from local intelligence about active nests.
LiPo fire response.
LiPo batteries can fail catastrophically — swelling, smoking, or igniting. The chemistry is different from regular fires; standard fire response is wrong. This section covers LiPo-specific response: prevention, recognition, containment, and what to do (and not do) if a LiPo ignites. Cohort default operations have not had a LiPo fire incident; the procedures here are precautionary.
LiPo failure stages — what to watch for:
Why standard fire response is wrong for LiPo
LiPo battery fires are chemical reactions, not simple combustion. The pack contains its own oxidiser; smothering with a typical fire blanket doesn't starve it of oxygen the way it would with a wood fire. Water can help cool the surroundings but doesn't extinguish the cell reaction itself.
Specific things not to do:
- Don't use a CO₂ extinguisher — doesn't affect LiPo chemistry.
- Don't use a foam extinguisher — same; LiPo is self-oxidising.
- Don't pour water directly onto the burning pack — water won't stop the reaction and can cause electrical hazards if pack is connected.
- Don't try to move a burning pack with bare hands — can cause severe burns and ignite clothing.
The right approach: let the LiPo burn out in place; cool surrounding materials to prevent fire spread. Sand is the most reliable cooling agent; large amounts of water can work for surrounding objects but not the LiPo itself.
Prevention is the main defence. LiPo fires are rare in cohort default operations because the prevention is simple:
- Always charge in fireproof bags. Cohort default: ~₱600 LiPo bag; replace every 6 months or after any heat event.
- Never charge unattended overnight. Charge during the day when you're present.
- Never charge a damaged or swollen pack. Swollen packs go to disposal, not to charging.
- Storage at correct voltage (3.85V/cell). Storing at full charge for extended periods is the leading cause of swelling.
- Cycle tracking: retire packs from mission work after ~70-80 cycles or any heat/swelling event.
- Inspect every pack before use: look for swelling, soft spots, or unusual heat. Damaged packs go to disposal, not to flight.
The cohort field bag includes: one LiPo bag for charging, one separate "discharged" bag for spent packs, sand or fire blanket for emergency response. Total cost ~₱1,500. Worth every peso.
LiPo disposal
Damaged or end-of-life LiPo packs need correct disposal. The cohort default workflow:
- Discharge to zero: connect to a high-resistance load (a pair of 12V automotive bulbs work) until voltage is below 1V/cell.
- Submerge in salt water: 5L of water with 200g salt; submerge the discharged pack for 24+ hours. Chemistry stabilises.
- Dispose with general waste: cohort default is brought to a Davao recycling centre that accepts batteries; barangay disposal is acceptable for emergencies.
Don't throw an undischarged or damaged LiPo into general waste — it can ignite during transport or at the disposal site. The discharge-and-saltwater process makes the pack safe for normal disposal.
Post-incident protocols.
The immediate emergency is over. The drone is recovered, the injury is being treated, the LiPo is disposed of safely. The post-incident phase is about documentation, reporting, and learning. This section covers what to record, when to notify authorities, and how to ensure the incident becomes a learning event for the cohort and partner orgs.
The post-incident sequence:
When CAAP reporting is required
CAAP's reporting requirements for drone operations are evolving; this is a snapshot rather than authoritative guidance. Cohort default approach: when in doubt, report. Over-reporting is a minor administrative issue; under-reporting can be a serious regulatory problem.
Generally requires reporting:
- Injury to any person caused by drone operation (any severity)
- Significant property damage from drone (e.g., damage to a structure, vehicle)
- Airspace violation (entry into restricted zones, altitude violations)
- Drone lost or unrecovered
- Mid-air encounter with another aircraft (any type)
Generally does not require reporting:
- Self-only drone crash with no other involvement (still document internally; not CAAP-reportable)
- Minor LiPo swelling without fire
- Mission abort due to weather or equipment with safe landing
For specific guidance, see safety.html and CAAP's current circulars. Cohort engineering is happy to advise on grey-area cases via the graduates Slack workspace.
The cohort incident record. Cohort engineering maintains an anonymised incident log accessible to current cohort members and graduated graduates. The log includes:
- Incident type and category (cross-referenced to this page's sections)
- Brief description of what happened
- Response taken and outcomes
- Lessons learned and adjustments to cohort default procedures
The record is the cohort's institutional memory. It's what makes individual incidents — which would otherwise be private learning experiences — into program-wide improvements. Contributing your incident report is one of the most valuable things you can do for the cohort's long-term safety.
Personal incident log. Beyond the cohort record, graduates typically maintain their own incident log: every situation they considered notable, including near-misses that didn't become incidents. Reviewing this log periodically (monthly, quarterly) reveals patterns: types of conditions that lead to issues, equipment behaviors to watch for, recurring decision points. Personal pattern recognition is the deepest form of learning.
Emotional processing. Serious incidents — especially those involving injury or near-injury — affect the operator emotionally beyond the immediate response. Cohort experience is that:
- The first 24-48 hours are often shock; technical learning is limited.
- The first week brings replay, doubt, and sometimes avoidance of flying.
- 2-4 weeks brings calmer reflection and useful pattern extraction.
- 3-6 months brings full integration into the operator's mental model.
Allowing this processing time is part of professional emergency response. Pushing back into operations too quickly often produces second incidents; taking time for processing produces better operators long-term. Cohort culture explicitly supports time off after serious incidents; partner orgs are encouraged to do the same. Alumni Slack has channels for incident discussion among trusted peers, which many graduates find helpful.