Search

Comparisons and Basic Set-Up for Open-Source AID Systems

Open-source AID systems are reshaping diabetes care. Here’s what every clinician needs to know to support patients safely and effectively.

Written By: ADCES member Tavia Vital, BSN, BA, RN, CDCES, Director of Intensive Management, Directora de Servicios en Espanol Integrated Diabetes Services. Edited by ADCES & danatech clinical staff.

September 19, 2025


Important Note: Open-source AID systems are not FDA-approved; clinicians do not prescribe them but can support patients in safe and informed use. 

Where To Start

When someone with diabetes is considering an Open-Source Automated Insulin Delivery (AID) System, the first step is to help them determine if Open-Source is right for them.  The next step is to provide them with information that will help them select the system that is best suited to their individual needs. This decision should be based on their comprehensive understanding of individual self-management goals, strengths, challenges, available resources, and their comfort level with using a class of tools that offer a wide range of features.
Begin by having them assess key factors such as the type of cellphone they use, as well as the model(s) of continuous glucose monitors (CGM) and insulin pumps available regionally that are covered by insurance.  Through education on Open-Source AID systems and each individual system, the person with diabetes will be able to determine which algorithm type is preferred and which features will best support their diabetes management and lifestyle goals.  Next, have them consider which app’s features will best support their unique goals and challenges. 

 

Software Overview

Four main Open-Source AID Systems are available today: OpenAPS, AndroidAPS, Loop, and Trio.  Each of the options have ample documentation that includes an overview of the system, the system components, where the build code can be accessed, how to build your own app (or rig), tips, instructions for set up and use, and support resources.  

To access the code required to install the algorithm on the devices the individual with diabetes will need to access The Documentation, or The Docs of the Open-Source AID System that they have chosen. 

This can be found at the links below for each of the systems:

  • AndroidAPS: https://androidaps.readthedocs.io/en/latest/
  • Loop: https://loopkit.github.io/loopdocs/, https://www.loopnlearn.org/, and https://loopkit.github.io/looptips/
  • Trio: https://trio docs.org/ 
  • OpenAPS: https://openaps.readthedocs.io/en/latest/index.html
While a handful of other Open Source options exist (most are not thoroughly tested or kept up-to-date), the rest of this article focuses on the “Big 3” that are kept “current” with ongoing contributions from a skilled team of coders and developers: AndroidAPS, Loop, and Trio.

 

Below is a summary of  the “big three” Open Source Options

AndroidAPS (AAPS)

Core Algorithm: Oref     Type of Closed Loop:  Hybrid to full

Where the algorithm “lives”: Android Phone

Compatible Pumps: Bluetooth connections: AccuChek Combo, AccuChek Insight; DanaR,RS, or Dana-i; Diaconn G8; EOPatch2; Omnipod DASH; Medtrum Nano or 300U; Equil 5.3;

Additional connection device needed: Older Medtronic pumps with certain firmware†; OmniPod Eros

Required Additional Hardware: (only for Medtronic or Omnipod Eros) RileyLink, EmaLink, OrangeLink, DiaLink, LoopLink

Compatible CGMS: See the CGM with Available BG Sources list on AAPS website

Compatible Wearables:  A variety of Android Wear OS watches

 

Loop

Core Algorithm: Loop   

Type of Closed Loop:  Hybrid    

Where the algorithm “lives”: iPhone

Compatible Pumps: Older Older Medtronic pumps with certain firmware; OmniPod Eros, Omnipod DASH, Dana-i and DanaRS-v3, Medtrum (in testing only)

Required Additional Hardware:  RileyLink, EmaLink, OrangeLink, or similar (only for Medtronic or Omnipod Eros)

Compatible CGMS: Dexcom G5, G6/ONE, or G7/ONE+; Medtronic Enlite; some Libre sensors. See LoopDocs, and for all possible options.

 

Trio

Core Algorithm: Oref    

Type of Closed Loop:  Hybrid to full

Where the algorithm “lives”:  iPhone

Compatible Phone: iPhone

Compatible Pumps: Older Medtronic pumps with certain firmware, Omnipod Eros, Omnipod DASH; Dana-i and DanaRS-v3, Medtrum (in testing only)

Required Additional Hardware: RileyLink (only for Medtronic or Omnipod Eros)

Compatible CGMS: Dexcom G5, G6/ONE, G7/ONE+; Medtronic Enlite; xDrip4iOS, some Libre sensors. See TrioDocs for all possible options.

 

System Hardware

Compatible pumps communicate to the app through either Bluetooth or an additional radio-link device, depending on the pump. While all Open-Source Code options can utilize older Medtronic pumps with specific software versions, some of the systems can also link to Omnipod Eros or Omnipod Dash pods.  As of September 2025, certain models of Dana pumps (available in Europe, the UK, and Korea) are compatible with Loop and Trio.  AndroidAPS has several other pump types that can be utilized as referenced in The Clinician's Guide to Understanding Open-Source AID Systems.

There are a variety of compatible CGM options for each open-source app, including specific versions of Dexcom and Libre listed in The Clinician's Guide to Understanding Open-Source AID Systems. Additional “outside of the box,” apps may be able to link a non-compatible sensor to the Open-Source algorithm’s app. 

Algorithms

Open-source systems have a base algorithm at the core of what drives the system decision making process. There are two main algorithms used by open-source systems: OpenAPS Reference Design versions Zero and Onec (abbreviated as oref0 / oref1) and Loop. 

Though the base algorithm may be the same between some of the open-source systems the appearance, usability features, tools, and setup processes differ.  

  • Oref0 is referred to as Temp Basal looping.  
  • Oref1 uses the Temp Basal algorithm and includes Super Micro Boluses (automatic, small, boluses) which allows the algorithm to work more aggressively compared to the Temp Basal only option.

AAPS and Trio include the option to enable Unannounced Meal Boluses which respond to rapid rises in glucose typically caused by missed food boluses or under-dosing.  The “Unannounced Meal Boluses” feature can be used (optionally) to lessen the need for meal announcements… particularly those that are low in glycemic index/load.

To learn more about the math behind the algorithms, review The Docs:

Build Methods:

For Android APS, Loop, or Trio, any computer with internet connection (MAC or Windows) can be used to build the app.  

The instructions found in The Docs for each platform read like a recipe.  Attention to detail is needed: Each step must be followed carefully and in the proper sequence.  It is important to note as a provider, that this is the part of the process where many people feel overwhelmed or intimidated.  Some people are able to navigate through the documentation and build their own app successfully.  Others opt to work with someone with build experience, or partner with a company that specializes in building Open Source Code AID apps, or by enlisting the assistance HCPs that provides stand-by coaching. 

Setup

All Open Source AID Systems require the user to input basal, carb ratio, insulin sensitivity, and glucose target settings. From there, each system requires a low or minimum glucose safety threshold and additional maximum insulin dose related settings.  Additionally, myriad feature-based settings and optional settings to either toggle on or off, decide to leave as default, or modify over time are available (note: this is more of an option in Trio than other systems). In addition, optional customizations or features under development are available to knowledgeable and experienced users. 

Companion Apps

There are several companion apps optionally available to help provide remote oversight of insulin delivery, glucose levels, and predictions. Some can also allow remote bolusing or activation of temporary overrides.  These companion apps are especially beneficial for parents of children or teens, or for adult caregivers of parents or other family members or close friends who may need oversight due to cognitive limitations.

Some companion apps provide additional features and benefits such as a variety of alarms that can more easily be heard during the nighttime, easy access to data and reports, inter-device and inter-app connectivity.

Examples include:

Loop Follow

Loop Caregiver

NightScout

xDrip+/xDrip4iOS

Sugarmate or Spike

AAPSClient  

Key Points

  • Open-Source AID systems have more adjustable features and allow for greater customization than commercially-available systems.
  • The three most common open-source systems, AAPS, Loop and Trio, each has unique capabilities and is compatible with a number of pumps and CGMs.
  • Some systems have the ability to manage certain types of meals with minimal user intervention.
  • Remote monitoring of live insulin and glucose data is possible, along with remote control of insulin delivery.
  • Users have the option to build their open-source program independently or with the assistance of skilled technicians and healthcare professionals.

Resources for Healthcare Professionals

OpenAPS

LoopDocs

Trio

Android APS

 

 

References

Braune, K., et al. (2019). Open-source AID systems for people with Type 1diabetes: Perspectives from healthcare professionals. The Lancet Diabetes & Endocrinology, 7(8), 681–683.   

Braune K, et al. (2022) Open-source automated insulin delivery: international consensus statement and practical guidance for health-care professionals. Lancet Diabetes Endocrinol. 2022 Jan;10(1):58-74. 

Kirk, A. D., et al. (2021). User-driven design of open-source AID systems: A qualitative study of motivations and outcomes. BMJ Open Diabetes Research & Care, 9 (1), e002280.

Lum, J. W., et al. (2021). Outcomes of patients with Type 1 diabetes using open-source AID systems. Diabetes Technology & Therapeutics, 23(3), 184–191

Bequette, B. W. (2013) Algorithms for a Closed-Loop Artificial Pancreas: The Case for Model Predictive Control. J Diabetes Sci Technol. 2013 Nov 1;7(6):1632–1643. doi: 10.1177/193229681300700624

M.W. Percival, Y. Wang, B. Grosman, E. Dassau, H. Zisser, L. Jovanovič, F.J. Doyle, Development of a multi-parametric model predictive control algorithm for insulin delivery in type 1 diabetes mellitus using clinical parameters. J of Process Control. 2011 Mar;21(3):391-404 https://doi.org/10.1016/j.jprocont.2010.10.003

Kang, S.L., Hwang, Y.N., Kwon, J.Y. et al. Effectiveness and safety of a model predictive control (MPC) algorithm for an artificial pancreas system in outpatients with type 1 diabetes (T1D): systematic review and meta-analysis. Diabetol Metab Syndr 14, 187 (2022). https://doi.org/10.1186/s13098-022-00962-2

Wu, Zekai et al. (2025) “Open-Source Versus Commercial Automated Insulin Delivery System for Type 1 Diabetes Management: A Prospective Observational Comparative Study from Canada.” Diabetes technology & therapeutics, 10.1089/dia.2024.0561. 18 Mar. 2025, doi:10.1089/dia.2024.0561

 

FAQs

1. How does someone with diabetes decide if an Open-Source Automated Insulin Delivery (AID) System is right for them?

The decision begins with an honest assessment of personal self-management goals, strengths, challenges, available resources, and comfort with using highly customizable tools. Education about what open-source systems can and cannot do is key to making an informed choice.

2. What factors should be considered when selecting an Open-Source AID system? 

Important considerations include:

  • The type of cellphone the person uses (iOS vs. Android)
  • Availability and insurance coverage of compatible insulin pumps and CGMs in their region
  • The algorithm type (e.g., Loop vs. oref0/oref1) that best matches their diabetes management style
  • App features that support their unique goals, challenges, and lifestyle

3.  How can healthcare professionals support this decision-making process?

Providers can guide individuals by:

  • Reviewing available device and app compatibility
  • Helping identify which features align with the person’s needs
  • Offering education about the range of open-source options
  • Supporting confidence in choosing a system that fits their daily routines and long-term goals

 

The Danatech Digest

Join 25,000+ professionals who get our latest diabetes technology updates

DISCLAIMERS: 

This site and its services do not constitute the practice of medical advice, diagnosis or treatment. Always talk to your diabetes care and education specialist or health care provider for diagnosis and treatment, including your specific medical needs. If you have or suspect that you have a medical problem or condition, please contact a qualified health care professional immediately. To find a diabetes care and education specialist near you, visit the ADCES finder tool.

ADCES and danatech curate product specifics and periodically review them for accuracy and relevance. As a result, the information may or may not be the most recent. We recommend visiting the manufacturer's website for the latest details if you have any questions.

danatech is Made Possible by Grant Funding and Industry Sponsorship