MIT-Mexico

Universidad Nacional Autónoma de México (National Autonomous University of Mexico)

Laboratorio Internacional de Investigación sobre el Genoma Humano (International Laboratory for Human Genome Research)

www.liigh.unam.mx

Querétaro, Mexico

Melanoma is the deadliest type of skin cancer. There is a specific melanoma subtype, known as acral melanoma (AM), which is the most common type of the disease in several countries in Latin America, Africa and Asia, including Mexico. AM has been understudied compared with other types of melanoma since it represents a small fraction of the melanoma cases in countries that traditionally undertake molecular cancer research, such as the US, the UK and Australia. Because of this, the causes of AM are still unknown, since it is unrelated to ultraviolet radiation exposure, and no effective treatment options are available for patients suffering this disease.

Our research group, based at the International Laboratory for Human Genome Research (LIIGH), at the National Autonomous University of Mexico (UNAM), has been focusing since 2018 on the molecular study of AM. We have formed research partnerships with institutions both within Mexico (e.g., the National Institute for Genomic Medicine and the National Cancer Institute of Mexico) and abroad (e.g., the National Cancer Institute of Brazil and the Wellcome Sanger Institute). As a result of these efforts, we have now generated several levels of data, including tumour genomic and transcriptomic sequencing, proteomics, genotyping, and a wealth of associated clinical data, of nearly 100 AM patients. In this project, we want to integrate these different levels of data in order to see if we can predict outcome data, such as tumour depth, ulceration status, or treatment response.

For this project, the interested student would have these tasks:
1) Assess the missingness of data, explore ways to alleviate these issues and maximise data utility
2) Test different ways of integrating these data to assess whether useful outcome information can be derived
3) Present their results both in a report and in a presentation to the rest of the research group at the end of the internship.

We can also adapt this project to the applicant's preferred skills and interests - we have data and opportunities in the wet lab as well. We hope that this project is of interest and are ready to host a student and make them feel welcome here in Mexico!

Instituto Politécnico Nacional (IPN)

Centro de Investigación en Computación (CIC, Center for computing research)

www.cic.ipn.mx

Mexico City

The Government of Mexico City has about 9,000 bicycles that are available to people for a small annual fee. There are about 200 parking stations for these bikes, where you can take one, ride it to another part of the city and leave the bike there. These vehicles need to be taken from parking stations that are almost full of bikes, to near-empty parking stations, using a set of about 20 trucks that transport up to 40 bikes each, and another 20 trucks that transport up to 20 bikes each.

The problem to be solved is to design a close-to-optimal strategy to move the trucks, seeking to maximize the availability of bikes to passengers. Care needs to be given to transport time, vehicle capacity, time of the day, and expected ideal number of bikes at each parking station.

We have large quantities of data about the daily movement of the bikes, dating back to several years. The data base is kept up to date.

The intern is expected (1) to provide a paper solution to this problem; (2) to carry out its programming to completion; (3) to test the solution with real data; and (4) to document the work performed during his(her) stay.

The intern should be fluent in at least two computer programming languages and have good domain of data base technology. Acquaintance with free software tools is useful; it is also useful to build its own software tools, or glue code, as needed. Large quantities (20 megabytes, say) of data are normally processed. Some knowledge of statistics and maximization is useful. I prefer, but do not require, some knowledge of machine learning tools.

Jaime Peña Studio

LIB - Laboratorio de Investigación con Materiales Biogénicos

jaimepenastudio.com

Bacalar, Mexico

Jaime Peña Studio is a collective of creative minds inspired by the essence and magic of nature and the multiverse, using these elements to guide the creation of architecture aligned with planetary energy. Formed as a collaborative family, we operate within a fractal system to share knowledge widely, nurturing young creative talent and offering tools to support their growth and learning. The studio is the culmination of 20 years of research into bamboo and natural materials, applied to architectural and engineering projects. Our team includes architects, computational designers, and interior architects who blend handcrafted techniques with advanced technologies, merging digital and manual design processes.

Our aim is to create architecture that harmonizes with its natural surroundings, crafting spaces that are truly aligned with our planet. Currently, our 10-person team is based in a holistic center in the Magical Town of Bacalar on the Riviera Maya, home to a stunning lake of seven colors. We are looking for people who are passionate about innovation and nature, with an investigative and proactive spirit. Our office is a place of research, knowledge exchange, and personal interaction. We are always aiming to grow as professionals and individuals, placing a high priority on human connections. Knowledge in Rhino, Grasshopper, and visual programming is a requirement.

During this period, students/interns will assist in developing lightweight, three-dimensional bamboo structures, exploring structural strategies through simulations and programming inspired by nature and fractal systems. They will also contribute to researching and creating roofing finishes made from natural materials for bamboo structures. The aim is to apply and test these insights and contributions in real-world projects undertaken by the studio

The student’s role involves researching and developing strategies for bamboo structural designs, with a focus on connection methods, optimization, and structural simplification. Additionally, they will explore and test potential waterproof materials for coverings. Students should be proactive and solution-oriented, with proficiency in digital modeling and programming tools like Rhino and Grasshopper; programming skills are an advantage. They should be collaborative, open to knowledge-sharing with the team, and adaptable in both digital and hands-on work, including performing computational simulations and material testing.

Centro de Investigación en Ciencia Aplicada y Tecnología Avanzada (CICATA) del Instituto Politecnico Nacional

www.cicataqro.ipn.mx/cicataqro/qro/cicata/index.html 

Querétaro, Mexico

This project focuses on the design, development, and testing of a contactless magnetic gear system—a cutting-edge alternative to traditional mechanical gears. The system utilizes magnetic fields to transfer torque without direct physical contact, offering numerous advantages such as reduced maintenance, quieter operation, and improved durability.

The primary objective is to create a scalable prototype capable of demonstrating the system's efficiency and reliability in device applications, such as renewable energy systems. Students will contribute to various aspects of the project, including:

1. Magnetic System Design
-Analyzing the interaction between permanent magnets and their spatial arrangement to optimize torque transfer.
-Using simulation tools to model magnetic field distributions and system performance.

2. Mechanical Integration
-Designing housings and support structures to ensure precise alignment and stability of the magnetic components.
-Exploring modular designs for adaptability to different torque and speed requirements.

3. Prototyping and Fabrication
-Building and assembling prototypes using 3D printing.
-Collaborating with vendors for sourcing high-performance materials such as rare-earth magnets and low-friction components.

4. Performance Testing and Optimization
-Setting up experiments to measure torque transfer efficiency, operational range, and thermal behavior.
-Refining the design based on test results to minimize losses and improve scalability.

5. Applications Research
-Analyze potential use in renewable energy prototypes, such as eolic or hydrokinetic systems.
-Evaluate the environmental benefits of the system in question in comparison to conventional gear systems.

Students will gain hands-on experience with advanced engineering techniques, computational simulations, and experimental testing. They will also have opportunities to enhance their problem-solving skills, collaborate in a multidisciplinary environment, and contribute to a project at the forefront of mechanical and magnetic innovation. This project is an excellent opportunity to explore emerging technologies in magnetic systems and contribute to developing innovative solutions for modern engineering challenges.

Ideally, you'll have a background in mechanical or electrical engineering, physics, or a related field. It'd be great if you're familiar with CAD software and simulation tools. Some understanding of electrical and magnetic principles. Strong analytical and communication skills

Tecnologico de Monterrey

Smart Electronics Research Group

tec.mx/en/research/innovation-in-smart-digital-technologies-and-infrastructure

Monterrey, Mexico

The lure of quantum computation stems from its potential to solve problems considered intractable with current, classical computational resources, by exploiting principles of superposition and entanglement to enable speedups for certain classes of problems. The realization of this capability is still in the research and development stages, however, and quantum technologies at present remain firmly embedded in the Noisy Intermediate-Scale Quantum (or NISQ) era. 

The NISQ era of quantum computing is defined by one of the main limiting factors in quantum computing: the presence of noise across the system, which introduces errors in the desired quantum computation. While quantum error correction is the leading candidate method for achieving full-scale, fault-tolerant quantum computation in the long-term future, the resource requirements of quantum error-correction (in terms of the number of physical qubits and their error rates) are currently prohibitively high. Specifically, true quantum error-correction mechanisms require several additional qubits to maintain the information of a single qubit – in an approximately 1000:1 proportion, typically. Despite many advances in this area, true error-corrected and fault-tolerant quantum computing is still estimated to be a decade or so away. Consequently, it is of intense current interest whether some form of quantum advantage, or quantum utility, can be achieved with NISQ devices, without full quantum error correction. This is also an area of interest for both Dell and ASU. 

Quantum error mitigation (QEM) has emerged over the past decade as a method for reducing errors in quantum computations. Instead of full quantum error-correction, QEM algorithms aim to reduce errors in quantum computation via classical post-processing of the noisy measurement outcomes. Recent work has identified striking resource requirements (in terms of the number of quantum circuits executions, or “shots”) for effective error mitigation; however, these results hold in the worst case, i.e., for classes of circuits that are highly entangling and spread quantum information rapidly among the qubits. Error mitigation techniques may still be efficient for certain more realistic, problem-specific classes of circuits, and developing efficient QEM algorithms remains an active area of research. 

Machine Learning (ML) has been used in several parts of the quantum computing pipeline to help improve the performance of quantum algorithms. In the area of transpilation, for instance, where the original algorithm is translated into another representation that satisfies the constraints of the target hardware, ML has been used to identify the best qubit mapping and operation ordering to avoid noise in the computation. In the area of quantum algorithm search, where ideal quantum circuits are searched in a combinatorial space, ML has been used to eliminate the majority of poor candidates without having to execute them on the target quantum machine.

Encouraging results in these other areas of quantum computing inspire the investigation into how ML can help mitigate errors inherent to quantum computations. We hypothesize that, by capturing characteristics of quantum algorithms and quantum machines, a ML model could learn to associate them with the occurrence and propagation of noise across the circuit and how it affects the output. By understanding those relationships, we further hypothesize that it is possible to learn how to “denoise” the result in classical post-processing. Indeed, supervised ML excels at finding relationships between data characteristics and observed outcomes. Recent advances in denoising processes, especially through the advent of the diffusion models, have also demonstrated incredible capabilities in image and audio domains.

This project aims to fill the gap related to the lack of an evaluation framework to compare QEM approaches and, more specifically, ML-QEM approaches. Our intent is to develop a set of tools that enable quantum researchers and software developers to assess the performance of (ML-)QEM in a holistic and consistent manner.

Our goal is to:
(1) understand the advantages/disadvantages of the various approaches of ML-QEM, particularly for certain types of problems, noise models, circuit structures, etc.; 
(2) based on the previous point, provide practical guidance regarding what ML-QEM techniques work best for given sets of problems, and why. 

Our assessment tools should be extensible and pluggable, so that new (ML-)QEM approaches and benchmark datasets can be easily added.

CICATA Queretaro - IPN

Queretaro, Mexico

www.cicataqro.ipn.mx

This project focuses on developing a predictive model for soil moisture using advanced deep-learning techniques and remote sensing data. The model will predict soil moisture for up to 72 hours before adapting to diverse soil types and environmental conditions. Historical data from satellite missions like CYGNSS, SMAP, GOES, and Sentinel, combined with in-situ measurements, will enhance prediction accuracy and achieve a coefficient of determination (R²) above 0.88. The model's performance will be validated across regions with varying climates, including the USA, Spain, and Australia, providing insights for agriculture, hydrology, and climate change mitigation​.

The intern will focus on developing and implementing machine learning models to predict soil moisture using remote sensing data from satellite missions such as CYGNSS, SMAP, GOES, and Sentinel. Their primary responsibilities will include preprocessing large datasets, designing and training predictive models, and validating results across diverse climates in regions like the USA, Spain, and Australia. By achieving accurate soil moisture predictions, the intern will directly contribute to advancing our organization's efforts in climate resilience, precision agriculture, and hydrology. This work will provide actionable insights for sustainable land and water management, aligning with our mission to mitigate climate impacts through innovative technology.

Required skills for this project include proficiency in Python programming and familiarity with machine learning frameworks such as TensorFlow or PyTorch. Students should have experience working with large datasets and basic knowledge of remote sensing or geospatial data analysis. Preferred skills include prior exposure to time series analysis, deep learning architectures, and satellite data processing. Strong problem-solving abilities, attention to detail, and an interest in climate resilience or environmental sciences will also be valuable for successful project participation.

CICATA Queretaro - IPN

Queretaro, Mexico

www.cicataqro.ipn.mx 

This project focuses on identifying and tracking tropical cyclones using Machine Learning and remote sensing techniques. Students will develop models capable of predicting hurricane trajectories, destinations, and intensities with a minimum three-day window and a mean track error of 140 km. The work integrates CYGNSS (Cyclone Global Navigation Satellite System) observations and the International Best Track Archive for Climate Stewardship (IBTrACS). The ultimate goal is to enhance early warning systems and disaster planning by modeling the societal impacts of hurricanes on critical infrastructures, inhabited areas, and sensitive ecosystems.

Using remote sensing data and historical hurricane archives, the intern will design and implement machine learning models to predict tropical cyclone trajectories, intensities, and societal impacts. Their tasks will include data preprocessing, feature engineering, and model validation to achieve high prediction accuracy with low mean track errors. By contributing to developing these predictive systems, the intern will support the organization’s mission to enhance disaster preparedness and resilience planning. This work will directly benefit communities by improving early warning systems and minimizing the impacts of hurricanes on infrastructure, populations, and ecosystems.

Required skills for this project include proficiency in Python programming, a strong understanding of machine learning techniques, and experience working with geospatial or remote sensing data. Students should also know data preprocessing, feature engineering, and model evaluation. Preferred skills include familiarity with time series analysis, deep learning frameworks such as TensorFlow or PyTorch, and handling large datasets from sources like CYGNSS and IBTrACS. Additionally, a background in environmental science, meteorology, or disaster management, strong problem-solving abilities, and attention to detail will be valuable for successful project participation.

FEMSA is a company headquartered in Monterrey, Mexico that creates economic and social value through companies and institutions and strives to be the best employer and neighbor to the communities in which it operates. It participates in the retail industry through a Proximity Division operating OXXO, a small-format store chain, and other related retail formats, and Proximity Europe which includes Valora, our European retail unit which operates convenience and foodvenience formats. In the retail industry it also participates through FEMSA Health, which includes drugstores and related activities and Digital@FEMSA, which includes Spin by OXXO and Spin Premia, among other digital financial services initiatives.

In the beverage industry, it participates through Coca-Cola FEMSA, the largest franchise bottler of Coca-Cola products in the world by volume. Across its business units, FEMSA has more than 392,000 employees in 17 countries. FEMSA is a member of the Dow Jones Sustainability MILA Pacific Alliance, the FTSE4Good Emerging Index and the Mexican Stock Exchange Sustainability Index: S&P/BMV Total México ESG, among other indexes that evaluate its sustainability performance.

Universidad Panamericana is a private nonprofit research university founded in 1967. It has three main campuses and 6 branch campuses in Aguascalientes, Guadalajara, and Mexico City. Universidad Panamericana currently offers 43 undergraduate programs and over 128 graduate programs, with 15,200 undergraduate and graduate students served by 2,657 faculty (38% with PhD) and more than 1,000 staff members.

The Applied Science and Advanced Technology Research Centre (IPN-CICATA Querétaro) is located in the city of Querétaro, in the state of Querétaro, Mexico. It is part of the National Polytechnic Institute and is a scientific and technological research centre designed to serve as a link between the scientific community and the productive sectors of goods and services, to accompany them and offer solutions to their development problems.

In order to achieve this objective, the IPN-CICATA Querétaro develops scientific and technological research programs with an interdisciplinary approach, as well as training human resources at the highest level, thus making a decisive contribution to strengthening the quality and national and international competitiveness of the Mexican productive apparatus.

In terms of research, the IPN-CICATA Querétaro has carried out a large number of projects with the economic support of the IPN, CONACYT and industry, generating patents, utility models, prototypes and various developments in its five different research areas: Image Analysis, Biotechnology, Mechatronics, Alternative Energies and Materials Processing and Manufacturing, which are linked to the economic activity of the region and the country. 

Currently, in the IPN-CICATA Querétaro, the Masters and Doctoral programs have been maintained in the National Quality Postgraduate Program (PNPC) of CONACYT, since its entry in 2007, currently its status is consolidated. It also has a specialization and three programs in its modality with industry.

From 2003, when the first two graduates of the Postgraduate Program in Advanced Technology graduated, until January 2023, 380 students have graduated, of which 101 are doctoral students, 268 master's students and 11 specialized students. The enrollment in this semester is 65 students.