Skip to content

Welcome to News from Loop and Learn

Master Your DIY Closed-Loop Insulin Delivery System

Published on 29 May 2026

Special Edition for ATTD 2026

In this edition:

Overview

The #WeAreNotWaiting movement is an incredible thing that continues to grow and grow in influence: on people living with diabetes, on the web of technology we rely on, and even on the companies that make the hardware. In this special edition of News from Loop and Learn, we present points of view and snippets from presentations and meetings at the recent Advanced Technologies & Treatments for Diabetes (ATTD), which was held in Barcelona, Spain on 11-14 March 2026.

We are extremely grateful to members of the Loop and Learn community who donated to the Nightscout Foundation and directly to fund developers' and other contributors' attendance at ATTD, and to #dedoc° global network of diabetes advocates, which funds "voices" from around the world to attend and contribute in big ways.

Our thanks to #dedoc° voices Miroslava Calegari, LnL team member; Theresa Hastings, Nightscout Foundation Director of Events and Community Development; and Batool Adawi, and to Magnus Hiis, diabetes nurse and open source contributor, for bringing us news from ATTD. This is a short summary. Much more information- very worth the read- can be found on Miroslava's blog, #type1dmaniac and some topics have been and will be explored further in other issues of this newsletter.

The #dedoc° Perspective

Advanced Technologies & Treatments for Diabetes (ATTD) is one of the world's leading conferences on diabetes tech, bringing together researchers, clinicians, industry, and people living with diabetes. For users of OS-AID, ATTD matters because this is where many of the technologies, algorithms, and ideas behind AID systems (commercial but Open source, too) are discussed, challenged, and shaped.

At ATTD 2026, the #dedoc° symposium once again reminded us how far the diabetes community has come. What began over 10 years ago as a bold idea—bringing people with lived diabetes experience into scientific conferences—has now become a true paradigm shift.

As highlighted by #dedoc° founder Bastian Hauck, this movement has helped transform diabetes care toward genuine patient-centered collaboration. Today, people living with diabetes are not just participants, but active contributors shaping research, technology, and care.

We at Loop and Learn are especially proud to have team members who have been supported through the dedoc° voices scholarship program, helping amplify these essential perspectives on global stages.

Key takeaway: The future of diabetes care is built with people living with diabetes, not just for them.

Thoughts from a Few Attendees of ATTD

"It's beautiful to see the progress toward fully closed-loop systems. I'm amazed by how the type 1 community keeps pushing for better tech. It was really inspiring to meet the Nightscout Foundation and dedoc crews, and to see all the patient-led innovation happening around the world, like the ATTD Diabetes Tech Tank. Let's keep learning from each other." -J. Co, Philippines

"Meeting the open source community at ATTD made one thing clear—behind every algorithm are people living this every day. The conversations were honest, unpacking both the promise and the limits of AI and commercial systems, especially with carb absorption never being straightforward. What I loved most was how naturally it all flowed—personal laughter mixed with deep dives into AI and open source algorithms, like a normal breakfast chat. It reminded me that real progress happens when lived experience and technology sit at the same table." -Asra, Saudi Arabia

"ATTD was a powerful reminder that behind every innovation, there are real stories, challenges, and hope. What inspired me most was the shared commitment to make life easier and safer for people living with diabetes." - Batool, Syria

News from Nightscout Foundation

The Nightscout Foundation, with support from our membership, #dedoc°, and Advanced Technologies & Treatments for Diabetes, sent two board members and five open-source developers to the conference. The time was spent connecting, collaborating, and building momentum for expanded research and development efforts within and beyond the #WeAreNotWaiting community. Discussions focused on important topics, among them stronger cooperative efforts, and increased opportunities to meet and work together in person. Bringing OS-AID to increasingly prominent conferences was part of those discussions as well.

A growing focus is the use of OS-AID to assist in multiple areas of women's health related to hormone regulation. Members of the team also met with all CGM companies present at ATTD to explore opportunities for OS-AID integration. Additional technical partnerships emerged that could build upon the existing data infrastructure and technical capabilities.

The event fostered valuable connections with community members, long-time collaborators, and research organizations from around the world. Sessions throughout the conference highlighted emerging technical innovations and new opportunities for the open-source diabetes community.

The Nightscout Foundation and the developers in attendance have quite the to-do list ahead of them after this successful conference. Stay tuned for updates throughout the year as the seeds planted at ATTD grow into impactful efforts across the open-source community!

Selected Presentation Summaries

Loop and Learn team member Miroslava Calegari attended ATTD online this year and reported in depth on several topics of interest on her blog, #type1dmaniac. Links to extended articles are included in the following summaries.

Sufyan Hussain (King's College London & Guy's and St Thomas Hospital, UK) presented "Lessons from Open Source Systems", highlighting collaborations with Katarina Braune, Rayhan Lal, and others in the open source community. Ten key concepts were extracted from a recent article in Diabetologia and were discussed in this talk. Here’s a quick mention of these concepts. Be sure to see this blog article for a fuller explanation.

  1. Open-source AID systems show that evidence-based, patient-driven innovation can become clinically accepted
  2. Interoperability = freedom of choice, remote control, and user ownership of data
  3. Open-source AID now has ethical and professional backing for clinical use
  4. Technology evolves faster than clinical adoption; opensource shows a faster path than regulated options
  5. Open-source AID provides transparency, helping clinicians understand how algorithms actually work
  6. More accurate insulin modeling improves decisionmaking, especially during activity or unexpected changes
  7. Fully closedloop is the goal— but current systems are still hybrid
  8. Movement toward fully closed-loop systems is gradual rather than binary, with stepwise progress toward greater automation
  9. One-size-fits-all AID does not work for everyone— personalization is key
  10. Open-source AID offers benefits— but access and usability remain barriers

Rayhan Lal (Stanford Medicine, Stanford University, USA) presented "Closed-Loop Systems: Future Challenges and Considerations", asserting that the technology is not the barrier to better diabetes care and outlining what the barriers are. He emphasized limited access to advanced diabetes technologies, unequal distribution of benefits, high costs, restrictive industry and regulatory structures, and lack of user control over devices and data. He also highlighted gaps between marketing claims and real-world performance, noting that current systems still struggle with exercise, stress, and changing insulin sensitivity. In addition, he pointed to physiological challenges such as variability in insulin absorption and the slow action of subcutaneous insulin. Overall, he argued that the greatest obstacles are tied to access, data ownership, economic inequality, and industry priorities rather than a lack of technical capability. Read more on this talk here.

"Artificial Intelligence (AI) for Carb Counting and Insulin Dosing: How Not to Do It" was presented by Tim Street (Diabettech, UK). Tim explained that while AI is increasingly being used to support meal insulin dosing and diabetes decision-making, it is not yet reliable enough to safely replace human judgment. Studies showed that AI systems frequently make significant carb-counting errors, often overestimating meals by amounts large enough to increase the risk of overdosing insulin. Although these tools can provide helpful support, they are designed to generate convincing responses rather than accurately interpret medical data, and they may overlook important context or produce misleading recommendations. It was emphasized that AI should remain "human-in-the-loop," with careful oversight, verified data sources, and clear safeguards. Looking ahead, Tim argued that safer diabetes AI will require hybrid systems with strong constraints and validation rather than standalone language models. More on this talk and topic can be found here and in a recent News from Loop and Learn article.

In his talk, "The Need for Interoperability", Johan Jendle (Orebro University, Sweden) emphasized that true interoperability—the ability for diabetes devices and systems from different manufacturers to seamlessly exchange and use data—remains one of the most important and unresolved challenges in modern diabetes care. While pumps, CGMs, pens, and other tools generate enormous amounts of valuable data, proprietary systems, inconsistent standards, and regulatory barriers often prevent that information from flowing freely between devices, users and healthcare systems. He also highlighted growing concerns around data ownership, privacy, consent, and the ethical use of patient data, particularly as artificial intelligence and big-data systems become more integrated into healthcare. Ultimately, he argued that the future of diabetes care depends on balancing openness and connectivity with strong protections for security, transparency, and patient control over data. Read more here.

Two sessions focused on fasting, a historically tricky topic when combined with diabetes: "Fasting During Ramadan with AID Systems", presented by Mohammed Al-Sofiani (King Saud University, Saudi Arabia) and "Twenty-Five Hour Fasting with Hybrid Closed-Loop", presented by Revital Nimri (Institute for Endocrinology and Diabetes, Israel). Fasting is a complicated topic, and one worth all of us learning about because even if we don't do it intentionally, we all find ourselves fasting at one time or another. Some quick takeaways are that, with careful planning and medical guidance, AID systems are making fasting safer and easier to manage and we always must remember that insulin is still needed during fasting, to stave off hyperglycemia and DKA.

Please read more about these fascinating and helpful presentations at #type1dmaniac's blog. Look for more on this extremely important topic in future issues of this newsletter.

Magnus Hiis, diabetes nurse at Ersta Hospital in Stockholm, Sweden, and open source contributor, shared his takeaways from ATTD in the article "From Gene Therapy to Age Discrimination – How Technology is Reshaping Diabetes Care", as well. It is written in Swedish and is translatable.

Booths of Interest

LUMIQ

At ATTD Ignite 2026—a startup pitch session showcasing early-stage innovations—the award winner, Beyond Diagnostics, caught our attention with a simple but powerful idea: Lumiq, a test (similar to a strip) that can tell you in minutes whether your insulin is still effective to use. It's still in development, but we're excited about the potential—and keeping our fingers crossed to see it in real life soon.

Enhance-d

Unlike most tools focused on glucose alone, Enhance-d focuses on context. It brings together data from CGMs (Dexcom, Libre), Nightscout, Tidepool, wearables (Garmin, Strava, Apple Health), and nutrition apps into one place to help make sense of how exercise, insulin, and glucose interact.

Syai

At ATTD 2026, Syai introduced their Tag sensor: a very small, lightweight CGM (~1.3 g) offering 14-day wear, real-time readings, and a reported MARD of ~8.1% without the need for calibration. Still early in its rollout, but with ambitions toward longer wear (21 days!). It's already AndroidAPS compatible. Could Syai become a future iOS OS-AID-compatible CGM option? Let's hope!

Insulin Saver

Traveling with insulin? InsulinSaver is a tiny "set-and-forget" temperature sensor you keep with your insulin—no app, no charging—that quietly tells you if heat or cold may have affected it, so you're not guessing when numbers don't make sense.