Our Work

In the U.S. education system, experiences of dehumanization are all too common, especially for marginalized students. Schools can rehumanize education by supporting positive relationships among studen…

In the U.S. education system, experiences of dehumanization are all too common, especially for marginalized students. Schools can rehumanize education by supporting positive relationships among students and staff and promoting equitable systems. This brief offers a continuous improvement framework—the Plan-Do-Study-Act cycle—with evidence-based practices to strengthen these critical relationships and systems. School communities can identify, implement, study, and sustain small shifts to transform their school into a place that honors everyone’s humanity.

A climatology-normalised CRPS framework that decomposes the predictability of environmental time series into endogenous (history) and exogenous (forcing) components, with a redundancy measure; applied…

A climatology-normalised CRPS framework that decomposes the predictability of environmental time series into endogenous (history) and exogenous (forcing) components, with a redundancy measure; applied to reservoir, streamflow, solar and wind forecasting.

Multiobjective grid optimisation using PyMoo and PowerFactory

Maqolada nofilologik yo'nalish talabalarining kasbiy-kommunikativ kompetentligini rivojlantirish masalasi ko'rib chiqiladi. Tadqiqotning maqsadi sun'iy intellekt vositalari va CLIL yondashuvini birlas…

Maqolada nofilologik yo'nalish talabalarining kasbiy-kommunikativ kompetentligini rivojlantirish masalasi ko'rib chiqiladi. Tadqiqotning maqsadi sun'iy intellekt vositalari va CLIL yondashuvini birlashtirishga asoslangan o'qitish metodologiyasini ishlab chiqishdir. Adabiyotlar qiyosiy tahlil qilinib, ikki yondashuvning kasbiy ingliz tili darsidagi vazifalari aniqlashtirildi. CLIL kasbiy mazmun va til maqsadlarini belgilashga, sun'iy intellekt esa materialni moslashtirish va muloqotni mashq qilishga xizmat qilishi taklif etiladi. Maqolaning ilmiy yangiligi ushbu vazifalarni besh bosqichli o'qitish modeli va mustaqil nutqiy faoliyatni baholash mezonlari bilan bog'lashdadir. Sun'iy intellekt yaratgan sohaviy ma'lumotni tekshirish va talabaning yakuniy ishini mustaqil baholash zarurligi asoslanadi. Taklif etilgan metodologiyaning samaradorligini aniqlash uchun keyinchalik pedagogik tajriba o'tkazish lozim.

This publication is co-authored by Alejandro Urrutia and Tonderai Dombo. Zimbabwe’s enduring economic, social and political challenges are almost always framed as requiring national-level soluti…

This publication is co-authored by Alejandro Urrutia and Tonderai Dombo. Zimbabwe’s enduring economic, social and political challenges are almost always framed as requiring national-level solutions, particularly elite negotiations and nationwide dialogue initiatives. Yet, recent national-level processes have consistently struggled to gain traction. This raises an important question: what if the problem is not an inability to develop effective national processes, but rather the assumption that solutions must originate at the national level? Across Zimbabwe, communities have long relied on dialogue as a practical tool for managing disputes, negotiating interests, and addressing shared challenges. Across villages, districts and wards, Zimbabwean community leaders frequently use dialogue to manage and resolve conflicts. Yet, local dialogue processes are not systematically connected to one another, nor are they meaningfully linked to national-level decision making. In that regard, the question we address in this paper is whether addressing this dual gap can offer a more effective pathway to national transformation. Intended primarily for Zimbabwean civil society and community-based organisations that are leading, supporting or participating in local dialogue initiatives and seeking to expand their influence beyond the community level, the paper examines 1) the promise of scaling local dialogue initiatives in Zimbabwe, 2) the different pathways for scaling, and 3) the potential risks.

The rapid development of aerial robotics and unmanned aerial vehicles (UAVs) has transformed them from niche radio-controlled toys into critically important components of modern cyber-physical systems…

The rapid development of aerial robotics and unmanned aerial vehicles (UAVs) has transformed them from niche radio-controlled toys into critically important components of modern cyber-physical systems. Modern autonomous drones are designed with an open modular architecture, allowing easy integration of various sensors, computing units, and computer vision systems. This adaptability has facilitated their mass adoption in the military sphere, logistics, infrastructure monitoring, and rescue operations. However, the transition from controlling individual drones to the concept of fully autonomous swarms, where hundreds of devices interact with each other without a human operator, drastically exacerbates the cybersecurity problem. Traditional remote-control systems, which historically relied on open 2.4 GHz or 5.8 GHz radio frequency channels, no longer meet security requirements. The integration of aircraft into urban and tactical airspace requires the implementation of reliable IP-based communication protocols, in particular Wi-Fi family technologies. They are able to provide not only command transmission but also high-resolution video exchange with minimal latency. At the same time, ensuring cryptographic protection of these channels faces a fundamental obstacle: strict size, weight, and power constraints. Onboard microcontrollers of drones have limited battery resources and computing power, making traditional "heavy" encryption algorithms inefficient or even impossible to deploy in the air. This article proposes a comprehensive analysis and innovative approaches to building secure aerial robotics networks. The work begins with an overview of modern communication architectures, including the evolution of telemetry protocols. Next, fundamental vulnerabilities of existing networks are investigated, and cyber-physical threats, such as satellite navigation coordinate spoofing and control hijacking, are analyzed. The main focus of the research is centered on developing a mathematically and energetically efficient encryption and authentication system based on elliptic curve algorithms, the TLS 1.3 transport protocol, and modern key generation and distribution standards.

Suplementary material for the study "Adaptive AI Support in Learning Design: Effects on Teachers with Different Levels of Collaborative Learning Knowledge" Contents: 01_DesignTask.pdf: Design task br…

Suplementary material for the study "Adaptive AI Support in Learning Design: Effects on Teachers with Different Levels of Collaborative Learning Knowledge" Contents: 01_DesignTask.pdf: Design task brief and instructions provided to participants. 02_Prompts_HK.pdf: Prompts used for the high-knowledge (HK) condition. 03_Prompts_LK.pdf: Prompts used for the limited-knowledge (LK) condition. 04_Prompts_General.pdf: Prompts used for the general AI assistance condition. 05_Pre Task Questionnaire_Subjective & Objective Test.pdf: Pre-task questionnaire including subjective and objective pedagogical knowledge measures used for participant classification. 06_Post Task Questionnaire.pdf: Post-task questionnaire administered after the design task. 07_Interview Questions.pdf: Interview questions used in follow-up interviews.

1. Gene annotation of the genome (Polistes dominula). In gff format2. GO annotation:Using the Excon pipeline with the Eggnogmapper setting run. nextflow run main.nf -profile singularity -bg -resu…

1. Gene annotation of the genome (Polistes dominula). In gff format2. GO annotation:Using the Excon pipeline with the Eggnogmapper setting run. nextflow run main.nf -profile singularity -bg -resume --input input_single.csv --custom_config myriad2.config --eggnog_data_dir /home/ucbtcdr/Scratch/GOTITS_excon/excon_updates_17Mar26/eggnog_data --run_eggnog --skip_cafe --outdir results