The Portuguese Company Addvolt, was one of the winners at the 2023 edition of the Portugal Ventures Awards, an initiative aimed at startups that stood out the most in the national entrepreneurial ecosystem in 2023.
Addvolt, a company focused on Building smart and sustainable transportation worldwide, was recognized in the Startup Industry & Technology category, a testament to its innovative contributions to the field
Founded in 2014 by four alumni from the Faculty of Engineering of the University of Porto (FEUP) and launched by UPTEC, the Company was one of the selected projects of the 2014 and first edition of the CMU Portugal initiative inRES – Entrepreneurship in Residence Program – a business acceleration program for entrepreneurial teams in the area of ICT.
Addvolt has developed a Plug-in Electric system – a groundbreaking technology already patented in several territories – that ensures the distribution of fresh and frozen products without using diesel. Additionally, it is emissions-free and operates with low noise due to the combination of a gas vehicle with the refrigeration system electrically powered by Addvolt.
The 2023 company’s distinction follows its success in 2022, where it was already awarded by Portugal Ventures, underscoring the company’s commitment to excellence and innovation.
The Portuguese Language Operations platform Unbabel, a CMU Portugal longtime partner and currently one of the Program’s affiliated partners, has secured $21 million in funding. This latest funding round was made possible through the support of prominent investors including Iberis Capital, GED Ventures Portugal, Point 72, Notion, ScaleVentures Partners, and Caixa Capital. The capital injection will be used to drive Unbabel’s global growth strategy and strengthen its position on the path to profitability.
Established in 2013, Unbabel’s Customer Service Solution allows modern enterprises to understand and be understood by their customers in dozens of languages by combining human expertise and artificial intelligence (AI). Powered by AI and refined by a global community of translators, Unbabel combines the speed and scale of machine translation with the authenticity that can come only from a native speaker.
Unbabel HQ is currently located in San Francisco, California, and has offices in Portugal, the United States, United Kingdom, Romania, the Philippines, Germany, Bulgaria, and Israel.
The company was founded by Vasco Pedro, a graduate from CMU’s Language Technology Institute (LTI) and CMU Portugal project post-doctoral fellow. André Martins, CMU Portugal alumni, is the company’s VP of Artificial Intelligence Research.
The company has also led one of CMU Portugal Large large-scale collaborative Projects MAIA (2020-2023), focused on developing a multilingual conversational platform supported by machine translation and dialogue systems, where AI agents assist human agents.
Vasco Pedro, co-founder and CEO explained in a public notice disclosed by the company that “From day one, we believed that language is the way to create a common understanding among cultures, businesses, and customers worldwide. In the past year, we announced two significant acquisitions for Unbabel – the German EVS Translations and the Israeli Bablic – and this funding is another step aligned with our acquisition strategy. Making Unbabel profitable will enhance its scalability and that is what we are focused on, now with the vote of confidence from this group of investors.”
CMU Portugal Ph.D. student in Engineering and Public Policy, Afonso Amaral, was distinguished with the best paper at the Babbage Industrial Innovation Policy Awards 2022 and received a $10K prize money for his work. Afonso had already won a “highly commendable work” distinction in the last edition but, in 2022secured the first prize.
“Winning this award in such an early phase of my career is motivating. I thought only professors could get these types of awards. It feels excellent to be still doing my Ph.D. and be already recognized at this level!”
The Dual Degree Ph.D. student at Instituto Superior Técnico and Carnegie Mellon’s Engineering and Public Policy Department co-authored the winning paper “National and Sub-National Policy for Domestic Manufacturing Flexibility: A Policy Framework to Incentivize Flexibility Based on Lessons from the COVID-19 Medical Supply Response”along with Nikhil Kalathil,hisPh.D. colleague at CMU.
“Nikhil Kalathil and I have similar research interests and share a CMU supervisor, Professor Erica Fuchs. Our collaboration comes almost as a natural thing. We look at similar problems in two different regions: I focus on Europe, and he focuses on the United States. We decided to combine our findings and think about how countries (and states) could make the best use of the full suite of federal and local policies to leverage their domestic industry. In a sense, what we have done is that we have combined several best practices across US states and EU Member-states, and we put them together in a policy framework focusing on investments in manufacturing flexibility”, Afonso Amaral.
By combining their expertise, the authors introduced a policy framework to support economic dynamism and manufacturing flexibility that uses the full suite of local, regional, and national policy mechanisms. The goal of this framework is to appropriately incentivize both pre-crisis and during-crisis investments in flexibility among firms of all sizes, taking advantage of the specific strengths and weaknesses of firms of different sizes. The study performs a five country case study to unpack the different policies implemented by Germany, Portugal, Spain, the Netherlands, and the United States of America to help domestic firms increase or pivot their production of medical supplies and equipment during COVID-19.
As a curiosity, the prize money will eventually be applied, according to Afonso, to another common goal: “We are currently writing a book together where we expand this and other ideas on how nations can best leverage their domestic industry. Maybe we can use this money for the publishing process? That would be a good idea!”
André Martins, a Portuguese researcher from Instituto de Telecomunicações (IT) and CMU Portugal alum,has won a European Research Council Consolidator grant (ERC CoG) of 2M€ to study artificial neural networks applied to natural language processing (NLP).
The ERC Consolidator funding will be applied to the DECOLLAGE project (DEep COgnition Learning for LAnguage GEneration), focused on finding solutions to some of the fundamental problems of NLP, using an innovative interdisciplinary methodology that brings together tools from artificial intelligence, sparse modeling, neuroscience, and cognitive sciences.
The existing deep learning models for natural language processing, while sometimes impressive, are often unreliable and even misleading: they do not generalize well to new domains, they do not exploit contextual information, they are poorly calibrated, and their memory is not traceable. These limitations stem from their monolithic architectures, suitable for perception but unsuitable for tasks that require higher-level cognition.
“This project is a step further in overcoming the limitations of current NLP technologies, making it possible for humans and machines to communicate effectively in natural language and to work collaboratively to solve increasingly harder problems”, André Martins
With the funding available under this new Grant, the research will focus on three main steps. First, using uncertainty and quality estimates as a guiding principle for controlled generation, combining this control mechanism with efficiently encoding contextual information and integrating multiple modalities. Second, developing sparse and structured models of long-term memory, with attention to descriptive representations, and third, devising new mathematical models for sparse communication (bridging the gap between discrete and continuous representations), supporting end-to-end differentiability and enabling a shared workspace where multiple modules and agents can communicate. These innovations will be applied to highly challenging generation tasks, including machine translation, dialogue, and open-ended generation.
This ERC award follows an early starting grant won by the researcher in 2017, worth 1.4 million euros, applied to the DeepSPIN project that allowed him to develop his research in deep learning and structured prediction to solve challenging tasks in natural language processing, including machine translation, quality estimation, and syntactic parsing.
André Martins was the first CMU Portugal Dual Degree Ph.D. Alumni in Language Technologies, having concluded his doctoral degree in 2012 granted by Técnico and Carnegie Mellon University (CMU). Since then, he has been highly involved with the Program initiatives, supervising two Ph.D. candidates and leading two CMU Portugal Large Scale projects (MAIA and GoLocal) in strict collaboration with teams at CMU. He is currently Vice President of AI Research at Unbabel and an Associate Professor at Instituto Superior Técnico (Técnico). He co-founded and co-organizes the Lisbon Machine Learning School (LxMLS), already in its 12th Edition, which counts with the support of the CMU Portugal Program, and is a Fellow of the ELLIS society, the pan-European AI network. With over 100 papers published, +6,000 citations, and an h-index of 37 in top-tier conferences and journals, André stands among the top 2% of the most-cited scientists in his scientific field by Elsevier.
Mahmoud Tavakoli, director of the “Soft and Printed Microelectronics Laboratory” at the Institute of Systems and Robotics (ISR) of the Faculty of Science and Technology of Universidade de Coimbra (FCTUC), has just won a prestigious European Research Council Consolidator grant (ERC) for five years. The researcher will receive 2.8M€ to study futuristic electronic circuits, a topic he has been working on for several years under the scope of projects funded through Carnegie Mellon Portugal (CMU Portugal).
The ERC Consolidator Grant will support the project “Liquid3D: 3D Printed, Bioinspired, Soft-Matter Electronics based on Liquid Metal Composites: Eco-Friendly, Resilient, Recyclable, and Repairable”, which is already underway and aims to provide design freedom to scientists, allowing them to print futuristic soft electronics and soft machines. «The idea is to make a transition from rigid, breakable, polluting and battery-dependent electronics to soft, resilient, recyclable, and self-powered electronics. In this context, the Liquid3D project will develop a series of innovative printable composites based on liquid metals to print 3D functional cells for sensing, stimulating, processing, and storing energy that is entangled in a distributed system and in a three-dimensional architecture», explains Mahmoud Tavakoli.
According to the researcher, “the most impressive thing about these systems is that they will allow a new level of bioinspiration in man-made devices, which is not yet possible”. For doing that, this project will develop novel printable biphasic composites based on liquid metal, and methods to print them, and finally procedures that allow recycling them.
“ERC projects fund high-risk, high-reward fundamental science. My goal is to redefine electronics and robotics», shares the researcher, concluding that he foresees a fundamental change in the materials used in electronics and robotics and how they will be made. “We’re talking about a wide range of soft, self-healing materials that can be printed together to permit a level of bioinspiration never seen before. This is due to a variety of new printable materials and printing technologies that will allow printing sensors, batteries, and electronic circuits to function side by side, like what we see in biology».
The Liquid3D project was launched this month of January and will permit the implementation of three new laboratories at FCTUC, namely the Laboratory of Soft and Printed Materials, which is intended to develop new materials for the next generation of electronics and robotics; the Digital Fabrication Laboratory, to create and validate technologies for additive manufacturing of the developed materials; and the Microscopy and Characterization Laboratory, to characterize electrical, mechanical and optical properties of the materials and systems produced.
Mahmoud Tavakoli manages a multi-disciplinary research team of Electrical, Chemical, Biomedical and Mechanical Engineers that combines expertise in nano materials, polymers, and liquid conductors for the applications of soft robotics, soft electronics, smart textile, smart plastics, and health monitoring. His research work at the “Soft and Printed Microelectronics Laboratory” from ISR Coimbra has been strongly supported by CMU Portugal Program and Fundação para a Ciência e a Tecnologia (FCT). Through this partnership, the FCTUC team has collaborated with the Soft Machines Lab from Carnegie Mellon University’s College of Engineering on three projects: Stretchtonics, one of CMU Portugal Entrepreneurial Research Initiatives; WoW, one of our Large Scale Collaborative Projects which is led by Glintt; and the Exploratory research project Exoskins. The research developed under this partnership includes stretchable electronics, wearable computing, wearable patient monitoring, printed sensors and electronics, digital health, and digital biomarkers.
A team of researchers at INESC TEC and University of Munich, including Carnegie Mellon Portugal (CMU Portugal) Ph.D. student Tamás Karácsony, tested an innovative solution to classify seizures, the main symptom of epilepsy, using infrared radar and 3D videos. Nature Scientific Reports recently published the results of this work, coordinated by Tamás Karácsony’s supervisor and CMU Portugal Scientific Director, João Paulo Cunha, researcher at INESC TEC and Professor at FEUP.
Despite a vast amount of video material available on seizure classification, studies on the subject are still rare and even more rare are approaches for automated, AI-supported solutions. This new study presents a new approach, which is the first to consider near-real time classification from two second samples, showing the feasibility of a system to support diagnosis and monitoring process (based on action recognition) that uses deep learning. This technique allows distinguishing between frontal and temporal lobes seizures (the two most common classes of epilepsy), or non-epileptic events.
Epilepsy is a chronic neurological disease that affects 1% of the world population, with seizures as one of the main symptoms – whose semiology is crucial to diagnose potential occurrences. The analysis of seizures is usually done using 2D video-EEG (electroencephalogram) at epilepsy monitoring units (EMUs), by specialized healthcare professionals. “During clinical diagnosis, the clinicians use these videos to visually recognize movements of interests defined by movement features (semiology)”, explained Tamás Karácsony, researcher at INESC TEC and CMU Portugal Ph.D. Student at the Faculty of Engineering of the University of Porto (FEUP).
However, the semiology assessment is limited by a high inter-rater variability among said professionals, and despite being promising, the automatic and semi-automatic approaches using computer vision still depend on considerable “human in the loop” effort. “A patient is usually monitored for several days, which has to be fully reviewed afterwards for the seizures. This requires a lot of time and effort from the clinical staff”, added the researcher.
To overcome this, the team of researchers has developed a deep learning-based approach for the automatic and near real-time classification of epileptic seizures. According to the paper’s first author, Tamás Karácsony, ” We present a new contribution inspired by the way experts analyze the semiology of seizures, taking into account not only the presence of specific movements of interest in different parts of the patients’ bodies, but also their dynamics and their biomechanical aspects, such as speed or acceleration patterns, or range of motion.”
The team turned to the world’s largest 3D video-EEG database, and extracted videos of 115 seizures, first developing a semi-specialised and automatic pre-processing algorithm to remove unnecessary environments from the videos. In practical terms, two image cropping methods are combined – depth and Mask R-CNN –, providing a clean scenario and, consequently, improving the extraction of relevant information from the available videos, minimizing unrelated variations, and improving the seizure classification process.
In a further explanation about the process used, Tamás explained “Our solution uses an action recognition approach with an intelligent 3D cropping of the scene to remove unrelated information, such as clinicians moving around the patients. By removing it, our method significantly improved classification performance”. According to the CMU Portugal Ph.D. candidate, this research has also proven the feasibility of our action-recognition approach to distinguish two classes of epilepsy and the non-epileptic class, with only two seconds of samples, making it useful for near-real-time monitoring. In addition, the solution we propose can be used in other datasets of 3D video for analysis and monitoring of seizures”
Therefore, in translating this knowledge to improved diagnosis and treatment, the approach serves two purposes: “it can be used for monitoring and alarms – which can alarm staff; or, if the approach is transferred to an ambulatory setting, a caregiver, when a seizure is ongoing, resulting in a faster response, which might decrease associated risks and Sudden Unexpected Death in Epilepsy (SUDEP). Without a near real-time approach this would not be feasible”, said Tamás Karácsony.
More research is required before this system can be implemented in clinical practice. Nevertheless, in the long-term the system is expected to benefit the clinicians, clinics, and patients. “With automated diagnosis support, the clinicians have to spend less time reviewing the videos, thus can treat more patients, hopefully make better decisions, which reduces associated costs (material and health) for clinics and society”, he concluded.
The research has been partially funded by Tamás Karácsony CMU Portugal/ Fundação para a Ciência e a Tecnologia (FCT) fellowship. Tamás is an ML researcher at INESCTEC with an MSc in Biomedical Engineering from the Technical University of Denmark and an MSc in Mechatronics from the Budapest University of Technology and Economics. His main research interests include Computer vision, Action recognition, Biomedical Applications of ML, Neuroengineering.
His preliminary thesis title is “Interpretable DL Based Clinical MoCap for Epileptic Seizure Classification” and he will spend his research period at CMU supervised by Fernando De la Torre at the Carnegie Mellon’s School of Computer Science.
A team of researchers from Faculdade de Ciências e Tecnologia da Universidade de Coimbra (FCTUC) has developed a new technique to produce microchip integrated stretchable circuits, in a low cost and scalable way. The new technology opens door to build biostickers to monitor patient’s health, and create electronic textile for smart garments for athlete performance monitoring, or for fashion.
According to Mahmoud Tavakoli, the project Lead researcher and Director of the Soft and Printed Microelectronic Lab at ISR Coimbra, integrating microchips into the printed circuits in an efficient and cost-effective way, is the primary and most important challenge in the field of soft and printed electronics. Tavakoli states that “ In fact we developed a new soldering technique, that works for elastomeric circuits. Pol-Gel, is a simple technique for self-soldering, self-encapsulation, and self-healing, that allows low-cost, scalable, and rapid fabrication of hybrid microchip-integrated ultra-stretchable circuits. After digitally printing the circuit and placing the microchips, we expose the circuits to a solvent vapor that allows the integration of this solid-state microchips into soft-matter and stretchable printed electronics. We addressed a problem, that is central for scalable fabrication, and commercialization of various products. We have successfully found a way to allow the rapid integration of microchips in ultra-stretchable hybrid circuits”
This solution is a major step to produce these circuits in a low cost and efficient way and will allow many applications developed by different research groups, to come out of the lab and pave a step towards commercialization as the process eliminates many fabrication steps. These printed circuits have already proven to be successful when applied in wearable biomonitoring and biostickers for health applications helping to monitor patients’ heart rate, muscular activities, body temperature, brain activity, or even emotions.
Equally, the textile industry can benefit from this since it will be possible to produce smart textiles with integrated microchips at a high scale. According to Mahmoud Tavakoli “we will now be able to integrate electronics into the next generation of smart garments, whether it is to monitor athlete’s performance or to map kinematics of an actress, or merely for the next generation of modern fashion, in which textile can be used as a communication tool. When it comes to the existing production lines for flexible electronics, we expect some of them will replace their current soldering technique with this novel technique”
This technology is already patented by Universidade de Coimbra and Carnegie Mellon University. Along with the technology transfer office of the University of Coimbra, UC Business, the team is now looking for commercial partners to commercialize the solution in different fields of application.
Useful links:
Paper in Nature Communications: “Reversible polymer-gel transition for ultra-stretchable chip-integrated circuits through self-soldering and self-coating and self-healing” I August 2021 I Pedro Alhais Lopes, Bruno C. Santos, Anibal T. de Almeida & Mahmoud Tavakoli
The research was led by Luís Caires, CMU Portugal Scientific Director and Faculty member at Nova Lincs/ FCT-UNL and Frank Pfenning, CMU School of Computer Science, as a key component of Bernardo Toninho’s Dual Degree Ph.D. in Computer Science, supervised by both faculty. Toninho joined the CMU Portugal Program in the end of 2009 as a dual PhD student and graduated in 2015. His Ph.D. research was focused on the logical foundations of message-passing concurrent computation, based on a so-called propositions-as-types correspondence between logic and programming languages.
The paper “Dependent session types via intuitionistic linear type theory” was published in the ACM SIGPLAN Symposium on Principles and Practices of Declarative Programming (PPDP) in 2011 and will be presented at this year’s conference in September. This award is given each year to a paper from the PPDP proceedings published 10 years earlier and is intended to “recognize the authors’ contribution to PPDP’s influence in the area of declarative programming.” The winning work introduced a novel notion of so-called dependent session types, a highly expressive specification and verification mechanism for message-passing concurrent programs, able to check for functional and communication correctness properties of programs without running them.
More about the INTERFACES project
The “INTERFACES project – Certified Interfaces for Integrity and Security in Extensible Web-based Applications” was carried out between 2009 and 2012 by teams from Universidade Nova de Lisboa (UNL), Faculdade de Ciências da Universidade de Lisboa (FCUL), Carnegie Mellon University and the Company, OutSystems
The project targeted the development of new techniques and tools for enforcing security, integrity, and correctness requirements on distributed extensible web-based applications by introducing novel, semantically rich notions of interface description languages, based on advanced type systems and logics.
Key outputs of the INTERFACES approach included core typed programming languages and environments for building extensible certified web applications, as well as design and implementations of prototypes for specification, programming, and reasoning about case studies, in collaboration with the industrial partner OutSystems SA, developer of the Agile Platform, a widely used web-based application development environment at the time.
Manuela Veloso was ranked in the top 10 of the 35 women on the Academic Influence list which includes astronauts, founders and CEOs of well-known technology and researchers from around the world. The Portuguese researcher has been strongly involved with the Program since it was launched in 2006. She has been leading multiple collaborative projects between Portuguese and Carnegie Mellon research teams and is an advisor for dual-degree Ph.D. candidates mostly in Computer Science, Robotics and Machine Learning.
Manuela Veloso is internationally renowned for her work in artificial intelligence and being one of the world’s top computer scientists and roboticists. Veloso is now the Emeritus Herbert A. Simon University Professor in the CMU School of Computer Science and head of J.P. Morgan AI Research. After receiving this distinction, she revealed in a news article by CMU School of Computer Science that she was “glad to see CMU on the list. She felt like a role model to women in engineering and noted that women on the list challenged conventional thinking, changed systems and went against what was popular at the time or expected of them.”
She has received several academic awards during her career, has published more than 250 scientific articles and created the CoBots. She was president of the Association for the Advancement of Artificial Intelligence (AAAI) and founder of the Coral Research Laboratory. She is also responsible for RoboCup, an annual robot soccer championship, which had the first edition in 1997 in Japan.
This solution is a major step to produce these circuits in a low cost and efficient way and will allow many applications developed by different research groups, to come out of the lab and pave a step towards commercialization as the process eliminates many fabrication steps. These printed circuits have already proven to be successful when applied in wearable biomonitoring and biostickers for health applications helping to monitor patients’ heart rate, muscular activities, body temperature, brain activity, or even emotions.
Manuela Veloso was ranked in the top 10 of the 35 women on the 
