OpenAPS (Open Artificial Pancreas System) is a community‑driven, open‑source technology that enables people with diabetes to automate insulin delivery using a continuous glucose monitor (CGM), an insulin pump, and a small computer (like a Raspberry Pi or Intel Edison). While OpenAPS can dramatically improve time‑in‑range and reduce the burden of constant decision‑making, it remains a complex, DIY system that demands a thorough understanding of its components and potential failure modes. System failures and medical emergencies—whether from hardware glitches, software bugs, or user error—can escalate quickly. This guide provides an authoritative, practical framework for recognizing, preparing for, and responding effectively to these situations. By internalizing these strategies, you can maintain safety and confidence even when the system behaves unexpectedly.

Understanding Common System Failures

To respond effectively, you must first know where things can go wrong. OpenAPS failures typically fall into one of four categories: sensor malfunctions, pump issues, software or algorithm bugs, and connectivity problems. Each has distinct symptoms and remediation steps. Recognizing these categories early allows you to isolate the problem and take the right corrective action without unnecessary panic.

Sensor Malfunctions and Disconnections

The CGM is your system’s eyes. Common sensor failures include:

  • Dropped data: The sensor stops transmitting glucose values for extended periods (e.g., 15–30 minutes or more). Causes: sensor expiration, loss of adhesion, transmitter battery depletion, or interference from electromagnetic sources. If data gaps exceed 20 minutes, OpenAPS may revert to a default safety basal, which may not be ideal for your current needs.
  • Noisy readings: Erratic, implausible glucose values (rapid spikes or drops that don’t match fingerstick checks). This can result from pressure on the sensor (“compression lows”), dehydration, or a failing sensor. A single noisy reading can trigger an incorrect temporary basal rate.
  • Calibration errors: The CGM requests calibration but the fingerstick value is outside the expected range (e.g., more than 20% difference), or the sensor refuses to accept the calibration. Repeated calibration failures are a sign to replace the sensor.

Always verify dubious sensor readings with a fingerstick before acting on them. If the sensor consistently fails, switch to manual fingerstick monitoring and replace the sensor as soon as possible.

Insulin Pump Failures

The pump is the final actuator. Pump failures can stop insulin delivery entirely or deliver too much. Key issues:

  • Occlusion alarms: The pump detects a blockage in the infusion set. This stops insulin delivery until the set is changed. Even a partial occlusion can reduce flow and cause gradual hyperglycemia.
  • Battery depletion: If the pump battery dies, insulin stops. OpenAPS cannot override this. Always carry spare batteries and track usage—most pumps last 1–3 weeks on a single battery depending on type and usage.
  • Reservoir or cartridge issues: Air bubbles, cracks, or leaks can cause under‑ or over‑delivery. Air in the tubing can prevent insulin from reaching you, leading to high blood sugar.
  • Pump motor or electronics failure: Rare but critical – the pump may become completely unresponsive. You must be prepared to switch to injections or a backup pump.

Software Glitches and Algorithm Bugs

OpenAPS itself is software, and bugs happen. Symptoms include:

  • Core loop freezing: The rig stops calculating and issuing commands. Often signalled by a missing “loop” indicator in the interface or a stale timestamp. Reboot the rig to restart the loop.
  • Incorrect dose calculations: An edge case in the algorithm might cause a temporary high or low temp basal. This can happen if sensor data is corrupted or if the pump history is out of sync.
  • WiFi or Bluetooth communication errors: The rig cannot talk to the pump or CGM. This may show as “waiting for sensor” or “no pump communication”. Check physical distance and pairing. A simple restart often resolves transient errors.

Keep your OpenAPS software up to date and review release notes for known issues. The community frequently fixes edge cases that cause unsafe behavior.

Connectivity Issues Between Devices

OpenAPS relies on radio frequency (915 MHz or 868 MHz for the pump) and Bluetooth (for CGM and sometimes the phone). Interference, distance, or device pairing loss can break the chain.

  • Pump not responding to rig requests: Often caused by low radio range or battery. Move the rig closer to the pump.
  • CGM values not reaching the rig (e.g., Dexcom G6 transmitter out of range). The rig may show “waiting for sensor” for too long.
  • Phone app not receiving rig status (if using Nightscout): Check network connectivity and rig Wi-Fi settings.

Set up redundant communication paths where possible. For example, use a phone as a secondary display for CGM data.

Recognizing Emergency Situations

Not every failure is an emergency, but you must be able to differentiate quickly. Medical emergencies require immediate action, while technical failures may only need a reboot or reset. Always prioritize the person’s immediate health over troubleshooting the system.

Severe Hypoglycemia (Dangerously Low Blood Sugar)

Symptoms: confusion, inability to speak, seizure, loss of consciousness, or inability to swallow. Hypoglycemia below ~55 mg/dL (3.0 mmol/L) with neurological symptoms is a medical emergency. OpenAPS may suspend insulin, but if the CGM is inaccurate or the pump fails to suspend, you still need to act. If the person is unconscious, do not give anything by mouth—administer glucagon and call emergency services immediately.

Severe Hyperglycemia and Diabetic Ketoacidosis (DKA)

Symptoms: nausea, vomiting, abdominal pain, deep rapid breathing (Kussmaul respirations), fruity breath odor, and altered mental status. DKA can develop within hours if insulin delivery stops completely. Even with OpenAPS, a pump occlusion or site failure can lead to DKA if not caught promptly. Check blood or urine ketones whenever blood sugar stays above 250 mg/dL for more than two hours with no clear cause.

Complete Loss of Insulin Delivery

If the pump stops delivering (empty reservoir, dead battery, occlusion, pump failure), you have no automated safety net. This situation can lead to DKA in 4‑6 hours without intervention. Signs: blood sugar rising rapidly (often >20 mg/dL per 15 minutes), ketones positive in urine or blood. Immediate action: treat with manual injections or change the pump device.

Unusual or Unexplained Symptoms

Any sudden change in consciousness, severe headache, visual disturbances, or chest pain should be treated as a potential emergency, regardless of glucose reading. Trust your instincts – if something feels wrong, seek help. Do not assume it is diabetes-related; call emergency services if in doubt.

Preparedness and Response Strategies

Preparation is the most effective way to reduce harm. The following strategies should be part of every OpenAPS user’s routine. Build them into your daily habits so that emergency responses become automatic.

Maintain Manual Control Skills

You must be able to operate your insulin pump manually without the rig. Practice:

  • Setting a manual temp basal on the pump.
  • Delivering a correction bolus from the pump keypad.
  • Knowing how to disable the radio board (if your rig has a physical switch) or simply disconnect the rig from the pump by removing the battery or unplugging the cable.
  • Performing a manual site change without assistance from the loop.

Practice these skills monthly until they are second nature. Record a video on your phone showing each step—useful for you or a helper during a stressful moment.

Keep Backup Supplies Accessible

Create a “diabetes go‑bag” that includes:

  • Spare pump battery (ensure correct type, e.g., AAA or AA).
  • Extra insulin (vial or pen, not just what’s in the pump reservoir). Store at proper temperature.
  • Spare infusion set and reservoirs.
  • Lancet device and test strips (for fingerstick verification).
  • Fast‑acting glucose (glucose tabs, juice boxes, or gel).
  • Glucagon or dasiglucagon emergency kit (check expiration date).
  • Ketone test strips (urine or blood).
  • Spare CGM sensor and transmitter (if available).
  • Written instructions for your OpenAPS configuration (basal rates, ISF, targets) – paper backup in case phone dies.
  • Small notebook and pen for logging.

Keep this bag in a consistent location (e.g., your daily backpack, car, or bedside table). Tell a trusted person where it is.

Regularly Monitor Glucose Levels with Fingersticks

Relying solely on CGM is risky. Verify with a fingerstick at least when:

  • You feel symptoms that don’t match the CGM trend.
  • The CGM has been in a “sensor error” state for >20 minutes.
  • You are about to drive, exercise, or sleep.
  • You treat a low or high that seems resistant to correction.
  • After a site change or pump reconnection.

Set a rule for yourself: if you have two consecutive CGM readings that are more than 30 mg/dL apart from a fingerstick, switch to manual mode until the sensor recalibrates or is replaced.

Establish a Communication Plan

Ensure at least one family member, friend, or coworker knows:

  • How to recognize a diabetes emergency (especially signs of severe hypo/DKA).
  • Where you keep your backup supplies and glucagon.
  • How to contact your healthcare provider or emergency services.
  • That you use an automated insulin delivery system and how to disconnect it if necessary (e.g., turn off the rig).
  • How to use an emergency glucagon kit (practice with a trainer pen).

Keep a card in your wallet with these instructions and a list of emergency contacts.

Learn and Practice Emergency Protocols

Review the Official OpenAPS Emergency Guide and keep a printed copy with your go‑bag. Practice low and high treatments until they are automatic. For severe lows: administer glucagon and call 911. For severe highs with vomiting: go to the ER immediately. For pump failure: inject insulin manually using a pen or syringe based on your correction factor.

Steps to Take During an Emergency

When a real emergency unfolds, follow these action steps in order. Stay calm and rely on your training. Check the person’s consciousness and breathing first.

  1. Assess the situation quickly. Identify the most immediate threat: loss of consciousness? No insulin delivery? Severe hypoglycemia? If the person is unconscious, unresponsive, or seizing, skip straight to calling 911 and administering glucagon. If conscious, ask them what they feel and check CGM and fingerstick.
  2. Administer fast‑acting carbohydrates for hypoglycemia. If conscious and able to swallow, give 15 grams of fast‑acting carbs (e.g., 4 glucose tablets, 4 oz juice, or 1 tube of glucose gel). Recheck in 15 minutes. If no improvement or symptoms worsen, repeat and call 911. If unconscious, do not give anything by mouth – use glucagon injection or nasal spray.
  3. Seek medical help immediately for severe symptoms. Call 911 or your local emergency number if: you cannot awaken the person, they have a seizure, they have difficulty breathing, or vomiting prevents keeping down liquids. Also call if blood glucose remains above 400 mg/dL with moderate to large ketones for more than 4 hours despite manual corrections, or if the person shows signs of DKA (nausea, rapid breathing, confusion).
  4. Disable automated features if they are malfunctioning. If the rig is causing erratic or unsafe insulin delivery (e.g., aggressively stacking insulin or failing to suspend when needed), remove the rig’s battery or turn off the radio bridge. Switch to manual pump control until you can diagnose the issue. Never trust a failing system.
  5. Document the incident. After the acute danger has passed, write down the timeline: what was the CGM trend? What actions did OpenAPS take? What did you do? Include any unusual behavior from the pump or sensor. This documentation will be invaluable for post‑incident review and troubleshooting. Use the notebook in your go-bag.

Post‑Emergency Follow‑Up

Once the immediate crisis is over, take time to analyze what happened. This prevents recurrence and improves your system’s reliability. Do not skip this step—it is as important as the response itself.

Review the Data

Use Nightscout, OpenAPS logs, and pump history to examine the hours leading up to the failure. Look for patterns: did the sensor start dropping out after a certain activity? Did the pump occlusion occur after a specific site location? Understanding the root cause helps you decide whether to change hardware, adjust settings, or update software. Export logs for deeper analysis if needed.

Update Your Procedures

Based on the review, update your personal emergency plan. For example, if the failure was a dead pump battery, you might add a recurring calendar reminder to change pump battery every 2 weeks. If the failure was a software bug, consider updating to a newer version of OpenAPS or switching to a more stable algorithm like oref1. If a sensor repeatedly failed, try a different insertion site or sensor type.

Communicate with Your Healthcare Team

Share the incident with your endocrinologist or diabetes educator. They can help you adjust your backup protocol and may offer additional resources, such as a continuous ketone monitor option or a prescription for a backup pump. They can also document the event in your medical record for insurance purposes. Some clinics offer printed emergency action plans for DIY system users.

Practice a “Failure Mode Drill”

Schedule a monthly drill where you simulate a system failure (e.g., remove the rig battery and manage manually for 4 hours, or intentionally cause a pump occlusion to practice site change under time pressure). This builds muscle memory and confidence. Over time, you’ll react automatically instead of panicking. Include a second person in these drills so they learn too.

Building Resilience: Community and Continuous Learning

No one using OpenAPS is alone. The community has documented countless failures and solutions. Use these resources:

  • OpenAPS Documentation – comprehensive guides on troubleshooting and emergency procedures.
  • Loops & Bolts (a community site with practical tips for loopers).
  • JDRF – offers educational materials on hypoglycemia and DKA recognition.
  • American Diabetes Association – clinical guidelines for emergency care.
  • Tidepool Loop (if you are considering an alternative algorithm) – includes safety documentation.

Also, consider attending virtual or in‑person meetups (check the Looping community for events). Sharing failure stories helps everyone improve. Join a specific OpenAPS user group on Facebook or Discord to ask real-time questions. The collective experience of dozens of users can help you avoid pitfalls and refine your emergency protocols.

Conclusion

OpenAPS empowers people with diabetes by automating insulin delivery, but it cannot eliminate all risks. System failures and medical emergencies are part of the landscape. By understanding common failure modes, recognizing warning signs early, maintaining manual skills, keeping backup supplies, and practicing clear response protocols, you can handle these situations with competence and calm. No system is perfect, but a well‑prepared user is the best safety net. Invest in your preparation today—your health depends on it. Review this guide annually, update your go‑bag, and stay connected with the community to keep your skills sharp.