2019 @PT Calls for Exploratory Proposals

Scientific Area: Sustainable Cities

Abstract: Over the past years, a considerable volume of work has been carried out to assess urban risks resulting from individual natural hazards. However, assessing the probable cumulative impacts of multiple hazards has not yet entered the mainstream of research and urban management practice. A better understanding of the impacts of cascading and compound hazards, including future hazards arising from climate change, is a fundamental requirement to support more focused and participated engagement on policy actions by various stakeholders, ranging from multilateral agencies to local governments.

MIT-RSC aims at taking a step forward in filling this gap through the development of an integrated risk assessment framework for measuring impacts of multiple natural hazards in urban areas, based on the comprehensive analysis of their direct and indirect interrelations and consequences. Such a framework is aimed at constituting a useful decision-support tool, providing a singular standardized metric to measure aggregated urban risks, and to accurately investigate the potential impact of pre- and post-disaster strategies. Accordingly, this research will contribute to improving the protection of vulnerable communities to the impact of natural disasters. The long-term effects of this project will be the risk reduction of extreme natural hazards in urban areas, namely coastal and pluvial floods, earthquakes, landslides, and urban fires, and the improvement of the capacity of buildings and populations, making them more prepared and resilient. The developed framework will be tested and used by the Lisbon Metropolitan Region that will also disseminate it through other public and private operators in and outside of its edge, guaranteeing thus the impact of MIT-RSC.

PT PI:
Tiago Miguel dos Santos Ferreira – UMinho
Paulo José Brandão Barbosa Lourenço – UMinho

MIT PI:
John Ochsendorf – MIT Architecture

Scientific Area: Earth Systems: Oceans to Near Space | Climate Science & Climate Change

Abstract: Ocean oxygen loss is among the greatest impacts of climate change that will affect most marine biota including e.g. commercially important fish stocks. It is believed that between 1970-2010, oxygen loss happened at a rate of 0.5 to 3.3% from the ocean surface to 1000m [1]. In order to mitigate oxygen loss effects we need to better document and understand it.
While long-term ocean monitoring programs and mooring systems allow the study of physical and chemical water parameters for the last decades, we need paleoceanographic records for older time periods. 
Oxygen loss has higher impact at the ocean intermediate depths (200-1500m) and especially in areas where dissolved oxygen content is already low (<2 µmol/kg) by natural reasons, called oxygen minimum or deficient zones. The intermediate depths have been less frequently studied in spite of being keystone in important ocean physical and chemical processes, such as mode water formation, upwelling, mixing and stratification. These intermediate depths are also home of the diverse ecosystems engineered by calcifying cold-water corals (CWC) [2].

Besides their ecological relevance, CWC are also important geological archives for past ocean conditions. Their strong structures are less prone to remobilization at the seafloor and their aragonitic skeletons are datable with high accuracy [3,4]. It is also possible to use them to perform elemental analyses documenting several physical and chemical conditions (like temperature, carbonate chemistry, water mass tracing, ventilation, pollutants; e.g. [5,6]) on the same colony, producing contemporary results.   Chromium isotopes (delta-53Cr) have been investigated as a tracer for the paleoredox (past oxygenation) state of seawater, with a recent effort made in the investigation of marine carbonates as chromium archives [7–9]. Both foraminifera and tropical corals species investigated presented species-specific isotope fractionation during calcification [7,9] or post-depositional overprint [8] that needs further investigation. However, we believe that azooxanthellate CWC will be a better archive of Cr isotopes given the lack of interference from light and photosynthesis over the Cr system. 

This project will investigate the use of CWC Cr isotopes as tracers of ocean deficient zones in the eastern Atlantic intermediate depths, addressing two main questions:
1. Can we effectively determine delta-53Cr from cold-water corals?
2. and Does cold-water coral delta-53Cr reflect the local water d53Cr where they live?
If so, (3) Do corals from areas of different dissolved oxygen content show proportional differences in coral delta-53Cr? and (4) Do fossil corals reflect changes in past dissolved oxygen content? 

To answer these questions they will analyze Cr concentrations and isotopes from living CWC and compare it with delta-53Cr of closely located water samples. Two study areas were selected: the Iberia and the Namibia/Angola margins, two coastal upwelling areas which differ in dissolved oxygen content and in their upwelling intensity.
The Cr system is impacted by biological activity such as productivity and, thus, we will characterize the upwelling systems (using diatoms accumulation rate and assemblages’ composition as well as organic carbon content) to better understand any potential biologically related fractionation. 
From areas where water Cr concentration exist but not stable isotopes, we will use the existing linear relationship between delta-53Cr and log[Cr] based on the closed-system Rayleigh fractionation process [10] to compute the expected d53Cr of a region. If the proxy is validated, we will perform preliminary investigation on its paleo applicability by analyzing Mid-Holocene coral samples contrasted with the living samples.
If successful, this project will develop a new proxy to reconstruct the intermediate depth oxygenation state, which could improve the scientific community’s ability to understand and predict the location and behavior of oxygen deficient zones.

PT PI:
Lélia Maria Matos Branco – IPMA
Fátima Abrantes – IPMA

MIT PI:
Ed Boyle – MIT EAPS

Scientific Area: Digital Transformation in Manufacturing | Earth Systems: Oceans to Near Space

Abstract: We have witnessed the colossal development and progress of automation and AI in many sectors of our society such as industrial processes, robotics, autonomous systems, driverless cars, and many others. Agriculture, at different paces, has been embracing the new technological innovations and is becoming a sector of strategic significance because of the need to produce more food for an increasing population, under limited natural resources and climate change impacts. Precision agriculture (PA), a concept that can be understood as digital/intelligent or automated agriculture, encompasses the use of technology (software and hardware) in agricultural production, protection, monitoring, and management.

Due to the importance of agriculture in our lives, PA will play a major role in enhancing food security and safety, as well as environmental sustainability. Besides the potential for agricultural science, PA became very attractive for other scientific areas including robotics, satellite remote sensing, and AI/ML.
The interdisciplinary AI+Green project, which addresses some of the key challenges in PA, has the goal of increasing the precision and reliability of early detection and monitoring of pests in vineyards by developing a novel spatio-temporal data-information fusion system based on fundamental and applied techniques from remote (satellite) sensing, agriculture science, probabilistic machine learning, and aerial robots.    Despite the achievements in digital agriculture over the last decades, the agri-food sector is facing critical challenges worldwide. Increasing land degradation, driven by intensive monoculture practices, associated, for example, to the wide use of chemical products (e.g. fertilizers and pesticides) made from limited raw materials, is enhancing food production costs and leading to significant environmental damages (e.g. soil erosion and groundwater contamination). Due to the impact PA has in our modern lifestyle and on the environment, not to mention the societal and economic impacts, and its scientific-technical challenges, digital-robotic agriculture is, definitively, an exciting and pertinent research area. Although PA is recently making fast advancements, early warning systems for detection and identification of pests is an emergent research topic, with high impact in agriculture systems. It is well known that Portugal has a historical relationship with Agriculture, being a “living laboratory” for the development of novel solutions in the field and representing a scalable case study within Europe.

In this context, the AI+Green project’s big question is: Is it possible to efficiently combine and model satellite remote sensing and UAV data (ie, large-scale and local imagery) to obtain high-fidelity and reliable detection of pests in vineyard plantations, in order to improve its early fight and control, mitigating economic and environmental losses?

To answer this question and to formulate the problem properly, AI+Green has brought together researchers (from academia, stakeholders and SME) involved in interdisciplinary and complementary topics relevant to pursue an appropriate solution to this problem. The multidisciplinary nature of this research-exploratory proposal, its objectives and impacts, allows the project to have overlap with three of the MIT-Portugal themes: 1) Climate Science/Change; 2) Earth Systems: Oceans to Near Space; 3) Digital Transformation.
The first research-area is related to data, measurements and monitoring of the climate, atmosphere, near-space, weather, land, and other features. These topics are strongly aligned with the AI+Green project because it will make use of satellite imagery (via Copernicus Sentinel-2 data) in combination with climate, temperature, humidity and other features relevant for agriculture management.
The second theme involves, besides oceans, the near-space environment. Due to the use of realistic (ie, real-world) satellite data and information as part of the project tasks, this is a research theme directly aligned with the proposal.
Digital transformation is also a thematic area embraced by the project. More specifically, we will develop advanced algorithms, information fusion and AI-ML techniques towards a sustainable and digital agriculture.
Finally, AI+Green will consider a data-science integration concept. That means the project will develop a data-driven framework, allowing the development and public-sharing of relevant data, software, discoveries, tools, and reports.

PT PI:
Cristiano Premebida – ISR-UC
Gil Rito Gonçalves – Faculdade de Ciências e Tecnologia da Universidade de Coimbra

MIT PI:
John W. Fisher III – MIT CSAIL

Scientific Area: Digital Transformation in Manufacturing

Abstract: Many industrial operations involve the processing of particle-laden viscoelastic fluids (PLVF), in which the continuous or matrix phase possesses a viscoelastic character [SML, ACB]. Simulating the dynamical response of these systems is a current challenge of computational rheology, mainly due to the non-linear many-body interactions of the constituents, and the complex underlying rheology of the continuous phase [CF1, SK].
Several important industrial examples that emphasize the inherent challenges of PLVF can be found in numerous engineering applications [SML, ACB], e.g., polymer processing of highly-filled viscoelastic polymer melts, processing of semi-solid conductive flow battery slurries, cementing and hydraulic fracturing operations using solids-filled muds, as well as in biological application [DL, EJL], e.g., the flow-induced migration of circulating cancer cells in bio-polymeric media such as blood.   
Having in mind the optimization of parts or processes that incorporate “highly filled materials”, the detailed understanding of the fluid’s constitutive behavior is expected to have a major impact on the properties and performance.
With this aim in mind, computational rheology (CR) can play a key role in the optimization [SAF1, CF1-5]. However, the inherent complexity of PLVF systems hinders the development of appropriate CR codes. Such computational problems can be solved through Direct Numerical Simulations (DNS) [MM].
Due to limitations on computational capabilities, however, this method can only be used for cases where understanding the evolution in the configurations of a relatively small number of particles is sufficient. This, therefore excludes all of the industrially relevant applications mentioned above. For large-scale industrial applications, where it is only the bulk or ensemble-averaged behavior of the suspension mixture that is typically required, up-scaled three-dimensional numerical models based on the Eulerian-Lagrangian multiphase (e.g., CFD-DEM) formulations have been employed. When the matrix phase is a simple linear Newtonian fluid, these techniques have been shown to be promising candidates for fast predictive simulations [JZ, YCC, CF5]. However, when non-Newtonian effects in the fluid are present, then due to the complex nature of the PLVF continuum phase there is not in the literature any accurate expressions to rapidly evaluate the drag coefficient of particles translating in viscoelastic fluids, valid over a broad range of kinematic parameters.   
The first scientific contribution in this area is authored by the members of this project [SAF1], who employed three-dimensional DNS to parametrize the effects of particle volume fraction, fluid elasticity as well as the effects of fluid inertia on the drag coefficient of spherical particles translating in a viscoelastic-matrix fluid that is described by the linear Oldroyd-B rheological model [SAF1]. Although, the approximate drag closure model that we have developed is only valid for low Reynolds number flows, because we observed that inertial nonlinearities and the change in the wake structure behind the spherical particle disrupt the canonical behavior of the drag coefficient correction that we obtained at Re ≤ 1. To be able to apply it to practical relevant problems, this approach must be extended to higher dimensional formulations that are appropriate for more realistic viscoelastic fluid constitutive models, e.g., the Giesekus or FENE-P models. However, due to the complex behavior of the drag coefficient correction, no one was able to devise appropriate models.   

Due to the difficulty of these parametrization, this project aims at exploring the possibility of using Deep Learning (DL) techniques [YB] to disentangle the complex non-linear interactions, and develop expressions to evaluate the drag coefficient of particles translating in non-linear viscoelastic fluids. At the first step, the results obtained from DNS for different values of parameters are collected [EM]. This step is computationally expensive and should be done using high-performance computing facilities. Then, these results will be used to train a deep neural network to classify the calculated sequence data. For that purpose a long short-term memory (LSTM) network [SH] will be adopted. An LSTM network enables to input sequence data into a network, and make predictions based on the individual time steps of the sequence data.   

Concluding, this project presents a key step towards modeling the flow of PLVF. The long-term objective of the research work involved in this project is the development of a 3D numerical code able to cope with the transient dynamics of PLVF, based on the Eulerian-Lagrangian approaches, considering realistic rheological models and different particle geometries. The development of a 3D viscoelastic Eulerian-Lagrangian numerical model will surely contribute to a more complete understanding of many advanced manufacturing and industrial operations.

PT PI:
Célio Bruno Pinto Fernandes – UMinho
João Miguel de Amorim Novais da Costa Nóbrega – UMinho

MIT PI:
Gareth H. Mckinley – MIT MECHE

Scientific Area: Digital Transformation in Manufacturing | Sustainable Cities

Abstract: The European Commission reported that buildings are the single largest energy consumer in Europe and the overconsumption of non-renewable resources that leads to resource scarcity, the growing consumer’s requirements, climate change and the increasing need for rapid urbanization. Thus, it is important to establish more efficient construction methods that are able to overcome the lack of interoperability and productivity currently present in the construction sector.
Over the past few decades, building processes have been fundamentally based on manual labour, which is one of the reasons why the construction industry is regarded as low tech with low levels of innovation.  When compared to traditional methods like casting concrete in formwork, construction 3D printing (c3Dp) has revealed important economic, environmental and constructability advantages, such as reduction in building time and waste, mass customization and complex architectural shapes.  Furthermore, c3Dp is pointed out as a promising trend of the future in the Digital Transformation in industry, due to the potential association of 3D printing with Building Information Modelling (BIM) and artificial intelligence (AI).  
Overall, a lack of knowledge related to the thermal behaviour of additively manufactured building elements was verified, mainly due to the fact that research projects are at an early stage of application, focused on structural design and durability concerns. However, with the fast evolution of additive manufacturing methodologies for building construction, thermal insulation represents a knowledge gap that must be closed.  

In this exploratory project, a masonry block/panel with improved thermal behaviour will be developed using a construction 3D printing technology. It is intended to explore the thermal behaviour of additively manufactured building elements finding measures/solutions to improve it. For that, the incorporation of thermal insulation materials and different geometries will be evaluated. In a first stage the basic solution of masonry concrete block will be printed and tested in laboratory to evaluate its thermal properties. The experimental results will be used as inputs in a 2D/3D thermal simulation software. The same software will then be used to optimize the thermal behaviour of the block/panel, by changing the geometry and the materials of the basic solution. Finally, the optimized solution will be printed and new tests will be performed to validate the numerical results. 
After this important step, it will be possible to act in the mixture and develop a thermal block/panel with a single material or adding thermal insulation materials or additives to the 3D printed mix. To be more ambitious and considering the importance of a circular economy, the incorporation of waste materials in the block, acting as thermal insulation, will be analysed in order to invest in three important goals: Digital Transformation in Civil Industry, Energy Efficiency and Circular Economy.

PT PI:
Ana Sofia Guimarães Teixeira – FEUP
João Manuel do Paço Quesado Delgado – FEUP

MIT PI:
Christoph Reinhart – MIT Architecture

Scientific Area: Sustainable Cities

Abstract: This project introduces a disruptive vision where vehicles fuelled by pure ammonia (NH3) will be able to reach near-zero pollutant emissions in a highly efficient manner. This will break the current paradigms for sustainable cities mobility by enabling: 

• Full decarbonization system where vehicles rely on ammonia, which is an H2 energy carrier, and a carbon-free fuel.
• A clear advantage of an already established and reliable infrastructure for ammonia storage and distribution.
• A relatively easy fleet vehicle conversion. 

Our team foresees internal combustion engines (ICEs) fuelled with pure NH3 without the help of any combustion promoter as the most cost-effective energy bringer. Therefore, to prove the reliability of our vision, this project will develop a new strategy using ammonia itself as a combustion promotor to ensure high temperature and pressure operating conditions for 100% NH3 combustion.   

The main breakthrough of this project will be the proof-of-concept for a fully decarbonized transport system through an ICE fueled by pure NH3. The team will develop a first-time high-fidelity CFD model with accurate and detailed kinetic schemes validated under experimental conditions.
To accomplish such high-risk and high-gain breakthrough, this team will explore the following undertakings: 

• Detailed NH3 kinetic models developed under the RMG framework
• High-fidelity CFD NH3 combustion model coupled with uncertainty quantification including: a) optimal injection design [WP4.2] and b) NOx abatement by inner Selective Non-Catalytic Reduction (SNCR) targeting low emissions (< 100 ppm)
• Experimental data over flame characteristics and ICE operation as a Proof-of-Concept

PT PI:
Valter Bruno Reis e Silva – IPPortalegre
Mário Manuel Gonçalves Costa – Técnico

MIT PI:
William H. Green, Jr. – MIT Chemical Engineering

Scientific Area: Earth Systems: Oceans to Near Space

Abstract: The MIT Portugal Partnership 2030 (MPP2030) fosters research in four strategic areas, including “Earth Systems: Oceans to Near Space”. Analysis of samples collected from areas of the Earth covered with ice are fundamental to understand many of the current challenges and changes occuring in our planet, including the oceans, land and atmosphere. Ice from the surface of the Earth may also be used as planetary field analogue, as it has similar conditions to icy moons of Jupiter and Saturn. The icy moons of Europa, Ganymede and Enceladus are thought to have a subsurface ocean, with a potentially habitable environment that may host life. These icy moons are the ideal location to determine whether extraterrestrial life exists in the solar system. Its direct detection may be technologically challenging, but it is possible to indirectly detect life by the presence of biomarkers, either in the icy surface of these moons or encaged in salty icy grains that are ejected in plumes. These are the objectives of future space missions, such as the Europa Clipper from the National Aeronautics and Space Administration (NASA) and the JUICE mission from the European Space Agency (ESA).  

The main objective of the innovative “Spectroscopy detection of bio-signatures in natural ice samples as a proxy of icy moons” (BIOMOON) research plan presented in this proposal consists on determining the most suitable spectroscopic analytical methods that can be used in the search  for organic molecules that are representative of life forms (biomarkers) that may potentially exist in icy moons of Jupiter and Saturn. The exploratory project is then divided into four objectives:
(1) Literature review for the selection of key molecules and their optical properties; this will be achieved by identifying target compounds that can be used as signatures of extraterrestrial life.
(2) Spectroscopic analysis of the selected molecules using ultravioleta-visível (UV-VIS) and IV absorption/emission techniques in pure solutions and mixtures to investigate the best spectroscopic signals; this will be achieved by determining the spectroscopic properties of the selected compounds.
(3) Perform the same analysis but in artificial ice samples prepared in the laboratory that will act as analogues of the icy moons, using the known composition of icy moons of Jupiter and Saturn; this will be achieved by the identification of biomarkers on artificial ice samples.
(4) with the obtained results, perform similar analysis as in point 3) but this time in-situ analysis in the field in the canadian sub-Arctic with different types of ice, which can be potential analogues of the icy moons of Jupiter and Saturn.

The BIOMOON project will inform future life detection missions to solar system icy moons with a subsurface ocean. The exploration of ice in our solar system is recognised within the Global Exploration Roadmap (GER) of the International Space Exploration Coordination Group (ISECG) and ESA’s Space Exploration Strategy as a key step to explore the Solar System.

PT PI:
João Alfredo Vieira Canário –Técnico
Zita Martins – Técnico

MIT PI:
Dava Newman – MIT AeroAstro

Scientific Area: Earth Systems: Oceans to Near Space

Abstract: Portugal has the 10th largest Exclusive Economic Zone (EEZ) in the world, covering a vast region of the northeast Atlantic. Accordingly, the National Ocean Strategy presented for the period 2013-2020 emphasizes the strategic value of the Ocean as national and European priorities for scientific and economic activities. However, it is fundamental to promote technological innovation and investment in many areas of ocean research to promote the growth and competitiveness of the ocean economy, both national and internationally. Several EU programs for marine innovation indicate the need to improve the in-situ component of marine observation systems.
The major problems and constraints with present ocean technology are related to sensors being either too large and/or expensive, while requiring costly, non-rechargeable and/or low capacity batteries, and thereby, severely limiting immersion times and ocean data gathering capabilities.
In this context, it is fundamental to develop cost-effective and multifunctional platforms and sensors to provide reliable in-situ measurements of key parameters for long-term monitoring.  The solution of this problem must take advantage of a “new generation” technologies involving miniaturization; communication; energy storage; data acquisition, storage and transmission; standardization and automation. 
The recent development of self-powered buoys suggests its use for ocean monitoring in real-time, either by directly equipping the buoy with sensors or by powering autonomous vehicles such as AUVs and UAVs. Unfortunately, the current applicability of these buoys is still quite limited by their energy source. That is, existing systems often include solar panels and batteries whose lifetime and power output is insufficient for the needs of upcoming applications. 
Harnessing ocean wave energy with a reliable and autonomous converter able to operate in remote locations is, therefore, an appealing concept on how to power sensors and autonomous vehicles. This is especially relevant considering the high availability of wave energy in the Portuguese EEZ. 

The current project proposal aims to design and assemble a new purpose-built turbine-generator set to equip wave-powered monitoring buoys. This is the key component for electricity generation and storage that will enable continuous data acquisition under longer-term deployment periods at open sea.  The development of wave-powered monitoring buoys is not a new subject for IDMEC/IST.
The current proposal is in the scope of a research line initiated in 2015 with the mid-term objective of deploying a fully functioning device at open sea. In fact, the experimental data from buoy prototypes studied by IDMEC/IST will be used to design the turbine-generator set for it to be adapted to real operating conditions and not merely a theoretical hypothesis. 
The project shall deal with important technological challenges in both mechanical and electrical engineering. On the one hand, the main contribution from the scientific and design perspective is the development and validation of a multidisciplinary numerical model to couple buoys, wave energy turbines, electric generators, and batteries. On the other hand, the built prototype will enable the stakeholders of maritime industries access to a new source of electricity from renewable energy in remote maritime locations. The design under-study is based on a patented air turbine owned by IST. The literature review shows that this turbine has the highest efficiency among the self-rectifying turbines, tested in laboratory and sea, with results published in peer-reviewed journals.
Despite being proposed for large power generation for public grid supply, the success of the current project would open a new market for this turbine technology. The fact that most components are available commercially “off-the-shelf”, or easily accessible to the Portuguese and European industry production capabilities, will also facilitate the technology transfer between from IST to market. 
To conclude, it is important to highlight that the applicability of stand-alone wave energy converters is not limited to ocean monitoring. In fact, these devices are a remote and autonomous electric infrastructure capable to power future technologies yet to be invented.   

PT PI:
Luís Manuel de Carvalho Gato – Técnico
Joao Carlos de Campos Henriques – Técnico

MIT PI:
Paul D. Sclavounos – MIT MECHE