e-GUIDE

Electricity Growth and Use In Developing Economies

About

The Electricity Growth and Use in Developing Economies (e-GUIDE) Initiative seeks to transform the approaches used for planning and operations of electricity infrastructure in developing regions. We are constructing measurement and data analytics techniques that are scalable, transnational, and verified using real data on electricity consumption and infrastructure. We partner with electricity service companies to develop our techniques, deploy them at scale, and build capacity for data and analytics in the electricity sector.

Projects

Electricity Consumption Predictions. While there are multiple excellent integrated electricity masterplanning tools, all of them suffer from limited data on electricity consumption growth. Using a broad library of historic consumption data, satellite imagery and other big data sources, and custom deep learning models, we are creating an open service with API access to electricity consumption growth predictions for individual businesses and residences. Our work initially targets East Africa with a goal to extend throughout the continent and beyond to developing regions in South and Southeast Asia. Better data will enable scarce resources to go further.

Electricity Reliability from Satellite Data. Beyond access, the availability and predictability of reliability is crucial for community and economic growth. We are developing a novel technique to provide wide-area, long-term estimates of grid stability across sub-Saharan Africa and South and Southeast Asia using daily data on nighttime illumination. We are pairing these data with our consumption growth models and also will release the data publicly for other applications from the community.

Opportunities at the Electricity-Agriculture Nexus. In many low-income settings, electricity is not the only bottleneck hindering economic growth, and there is substantial latent capacity for economic growth in the agricultural sector. To help unlock this growth, we aim to build techniques to uncover opportunities where targeted and coordinated investments can simultaneously address agricultural bottlenecks as well as energy bottlenecks. In the long term, we intend for this tool to provide inputs to integrated energy planning tools on prospective agricultural investments and their effect on electricity consumption.

Outputs

People

The e-GUIDE Initiative is a partnership of groups at four universities: the University of Massachusetts - Amherst, Carnegie Mellon University, Columbia University, and Colorado School of Mines. The work of the Initiative would not be possible without the generous support of The Rockefeller Foundation.

Jay Taneja
STIMA Lab
UMass Amherst

Paulina Jaramillo
Green Design Institute
Carnegie Mellon

Nathan Williams
Exec. Director
Carnegie Mellon

Vijay Modi
QSEL
Columbia

Morgan Bazilian
Payne Institute
Colorado School of Mines

Chris Elvidge
Earth Observation Group
Colorado School of Mines

Kim Baugh
Earth Observation Group
Colorado School of Mines

Joel Mugyenyi
Research Fellow
e-GUIDE

Terry Conlon
PhD Student
Columbia
Interview

Santiago Correa
PhD Student
UMass Amherst
Interview

Simone Fobi
PhD Student
Columbia
Interview

Jorge Izar
PhD Student
Carnegie Mellon
Interview

June Lukuyu
PhD Student
UMass Amherst
Interview

Lefu Maqelepo
PhD Student
Rochester Institute of Technology
Interview

Aggrey Muhebwa
PhD Student
UMass Amherst
Interview

Bob Muhwezi
PhD Student
UMass Amherst
Interview

Zeal Shah
PhD Student
UMass Amherst
Interview

Contact

We'd love to hear from you with questions, comments, and interests in collaboration!

info@eguide.io

Interview with PhD Student - Zeal Shah
By Joel Mugyenyi


Q: Brief background about yourself/PhD program?
A: I hold a Bachelor’s degree in electrical engineering with specialization in power systems. Right after finishing my undergraduate studies, I moved to Pittsburgh to pursue a Master’s degree in Energy Science, Technology, and Policy at Carnegie Mellon University. After graduating from CMU, and before joining Prof. Taneja’s research group, I interned at a smart metering solutions firm called SparkMeter.

Q: Why did you decide to pursue a PhD/Research?
A: I decided to pursue a Ph.D. in order to learn and gain experience in using and applying computing tools to solve problems in the domain of power and energy. Furthermore, my work at SparkMeter and discussions with Professor Jay Taneja made me realize the importance of addressing the energy problems in developing economies. I joined Prof. Taneja’s group because it provided me with the right research platform to develop my computing skills and address various energy problems with a possibility of making an impact in the real-world.

Q: What does e-GUIDE mean to you?
A: According to me, e-GUIDE initiative is aimed at addressing one of the most important questions in electrification planning – “What should be the capacity and how flexible should the new electricity supply system be?”. e-GUIDE products will provide planners with an ability to study the growth in demand over time. This new piece of information will help planners in optimally sizing the power systems so that the future demand could be catered to without any loss of reliability. Additionally, demand growth information will help investors in developing better project and energy financing models.

Q: What do you think will be the most significant impact of e-GUIDE?
A: Innovative, affordable and practical tools to predict demand and monitor electric grid.

Q: Tell me about your e-GUIDE project?
A: I am working with a research team at Colorado School of Mines that is developing power systems reliability metrics using satellite imagery. Currently my work is focused on verifying the reliability metrics developed by the Mines team. Verification process involves studying the correlation between nightlights-based reliability measurements and variety of ground reliability measurement datasets like ESMI Kenya. Main goal of our work is to figure out the feasibility of monitoring grid reliability using satellite imagery because satellite imagery data is affordable, consistent and regularly updated.

Q: What’s the status of your e-GUIDE projects thus far?
A: My work is still in its nascent stage. Our initial results will be published in a poster paper at BuildSys’19.

Q: What are your materials and methods for accomplishing your project?
A: Just data analysis for now.

Q: What are your major challenges in carrying out these projects?
A: I am waiting to receive daily nighttime lights datasets from the Mine’s team. Apart from that I haven’t faced any challenges yet.

Q: Have you figured out how to overcome these challenges?
A: I am not very sure of the challenges that lie ahead and so I haven’t given it much thought, but yes, in case of computational challenges, Jay, Simone, and Aggrey will be the best resources for me.

Q: What do you think the impact of your specific project will be?
A: If we are able to verify and confirm the feasibility of using satellite imagery for monitoring grid reliability, we will be developing a one of its kind tool/API for wide-area grid reliability measurements. It will be a cost-effective solution for utilities in the developing countries to monitor their systems and focus their efforts and spend resources only on regions experiencing high unreliability. Additionally, international development partners will be able to monitor the systems independently.

Q: How do you think your project compliments other e-GUIDE projects?
A: My project is not directly related to the e-GUIDE objectives. My work is aimed at helping utilities in maintaining their existing systems better while e-GUIDE’s main objective is to help utilities in planning new systems/extending the old ones.

Q: What do you like most about your project?
A: I get to develop skills in working with large geospatial datasets.


Interview with PhD Student - Lefu Maqelepo
By Joel Mugyenyi


Q: Brief background about yourself/PhD program?
A: I grew up in rural Lesotho. I hold a bachelor’s degree in Mechanical Engineering from the University of Botswana and a Masters’ of Science Degree in Electrical and Computer Engineering from Carnegie Mellon University. I am currently a PhD student in Engineering and Public Policy at Carnegie Mellon University.

Q: Why did you decide to pursue a PhD/Research?
A: I am doing a PhD to address the energy poverty problem in Sub-Saharan Africa. SSA has a higher percentage of population without access to modern energy services, especially electricity. While this problem has most widely been looked at from a technical perspective, the fact is that it also must be looked at from a policy perspective. The policy perspective encompasses things like funding mechanisms, environmental sustainability, etc. I hope to contribute towards informing policy decisions with regards to increasing energy access.

Q: What does e-GUIDE mean to you?
A: The main objective of e-GUIDE as an initiative is to transform the approaches used for planning and operations of electricity infrastructure in developing regions. Personally, it is a platform that enables me to take part into what I believe will be high impact work, touching many lives in the process.

Q: What do you think will be the most significant impact of e-GUIDE?
A: The initiative will foremost 1) supplement the tools used by planning departments of various utilities across SSA and 2) help inform policy decisions by governments, utilities and all relevant stakeholders involved in providing energy access.

Q: Tell me about your e-GUIDE project?
A: The project title is ‘Assessment of subsidies for rural electrification in East Africa.’ The basic idea is to evaluate, using data from actual utilities, the subsidies that go into financing provision of energy access by extending the national grids. After evaluating these subsidies, we want to look at what impact the same subsides would have if applied to off-grid solutions in the form of micro-grids. The questions this research aims to answer are 1) is it possible to apply these subsidies to micro-grids, 2) if applicable, would they make the micro-grid space cost competitive to the grid from a customer and investor perspectives respectively and 3) what would be the ideal structure of these subsidies to create an optimal ecosystem ensuring sustainable, environment friendly and reliable electricity service?

Q: What’s the status of your e-GUIDE projects thus far?
A: This project is at the data collection state. We are collecting data from utilities in east Africa (mainly Rwanda thus far).

Q: What are your materials and methods for accomplishing your project?
A: The materials are the data from the utilities. These data include cost data of the generation, transmission and distribution infrastructure, consumption data, customer location data etc. As an overview, the correct evaluation of subsidies relies on knowledge of the true cost of service or product. For us, the service of interest is the delivery of energy to a community. We therefore do this by 1) evaluating what the true cost of the whole power system is, 2) allocating this cost to different communities served by the grid and 3) by finding the difference between the cost and revenue generated from these communities. Existence of subsidies are characterized by costs eclipsing the revenue.

Q: What are your major challenges in carrying out these projects?
A: The main challenge is getting the data at the level of detail that would make the analysis a bit straight forward. In most cases there are no records of data of interest, or when available, at times it is aggregated. Unavailability and aggregated data make it much more complicated to allocate the cost in a manner that is sensible and reflective of reality.

Q: Have you figured out how to overcome these challenges?
A: One of the ways of overcoming these challenges is to make a strong assumption that communities with similar features must behave in a similar fashion. This allows for translation of what happens in one community to another of similar characteristics.

Q: What do you think the impact of your specific project will be?
A: At its successful completion, the results of this project will culminate in a predictive simulation model that will help utilities estimate the potential need of subsidies in yet to be electrified communities. With this knowledge, they will then make a more informed decision as to how to best provide energy access to such communities.

Q: How do you think your project compliments other e-GUIDE projects?
A: This project relies heavily on being knowing what consumption levels are. For yet to be electrified communities, this will be unknown. This calls for better demand prediction, which is one of the e-GUIDE projects. To project into the future, there will be heavy reliance on results from demand growth prediction models from the other project.

Q: What do you like most about your project?
A: Besides its apparent importance, it is a novel approach from a quantitative perspective. It is excitingly challenging, and I learn more about one of the core backbones of any country’s economy, the power system. The fact that I do work that may potentially transform how the power system is planned in the future, for the better, is also very gratifying.


Interview with PhD Student - June Lukuyu
By Joel Mugyenyi

Q: Brief background about yourself/PhD program?
A: I am from Nairobi originally and I came to the US in 2009 for my undergraduate degree at Smith College and graduated in 2013. I then started working for a software company in Bedford called Aspen technology. While working there, I did a distance learning master’s program from Loughborough university in the UK focusing on Renewable Energy systems. I started my PhD program in September of 2018. My research at Smith College focused on analyses of energy systems in developing countries, mainly micro-grids, such as power flow analysis and over the years I have broadened my research interests to include economic and risk analysis of energy projects. For example, during my master’s degree I looked at the techno-economic feasibility and risk analysis of a community based milk cooling system in Tanzania.

Q: Why did you decide to pursue a PhD/Research?
A: I really enjoy research work. Starting from Smith, I concentrated on a research focused approach looking at problems and trying different ways to solve them. A PhD was the next step in my research journey

Q: What does e-GUIDE mean to you?
A: It’s a community of people working on the same kind of problems, that is, energy access challenges in the developing world. The team has a diverse set of skills with members all at different places in their careers, but we are all working towards the same objective. Therefore, through collaboration and knowledge sharing, we can work towards achieving our shared goal of increasing sustainable energy access in the developing world.

Q: What do you think will be the most significant impact of e-GUIDE?
A: I think the kind of things we are proposing to do have not been done yet in this space, for example, predicting consumption of customers who have not yet been connected. I feel e-GUIDE will have a huge impact in such areas.

Q: Tell me about your e-GUIDE project?
A: I am working on a project looking at opportunities for coordinated energy and agricultural investment. We are currently focusing on irrigation in Ethiopia and we want to use pollution data from a weather satellite to predict where a lot of diesel-powered irrigation pumping is occurring in Ethiopia so as to inform where investment in establishing or extending electricity infrastructure could be targeted.

Q: What’s the status of your e-GUIDE projects thus far?
A: I am just getting started with the project and right now it’s pretty much data cleaning and processing and early stages of analyzing the data.

Q: What are your materials and methods for accomplishing your project?
A: I have had to use computing clusters which I am new to. I will also need to use clustering


Interview with PhD Student - Aggrey Muhebwa
By Joel Mugyenyi

Q: Brief background about yourself/PhD program?
A: I am currently doing my PhD with the ECE department at the University of Massachusetts, Amherst. I have not yet narrowed down my concentration. My current work is aligned with the application of deep learning and computer vision techniques to satellite imagery to analyze infrastructure in the developing world. I am originally from Uganda.

Q: Why did you decide to pursue a PhD/Research?
A: I was looking for research opportunities with a focus on the developing world. I was personally motivated to engage in work geared towards developing socio-economic tools for improving accessibility to amenities such as electricity for people in the developing world.


Q: What does e-GUIDE mean to you?
A: To me, e-GUIDE means a group of researchers who have the opportunities and means to improve electricity accessibility in the developing world by leveraging computing tools and who, additionally are not constrained by bureaucracies that can be found at large utility organizations

Q: What do you think will be the most significant impact of e-GUIDE?
A: From my point of view, e-GUIDE has great partners in the developing world such as Kenya Power in Kenya, Umeme in Uganda, REG in Rwanda and thus the opportunity to influence decision makers therefore having a direct impact on the ongoing electricity accessibility projects. e-GUIDE has data and proof to help decision makers use their limited resources in ways that can most effectively improve people’s lives

Q: Tell me about your e-GUIDE project?
A: I definitely feel like I am still defining my project. The synopsis of the project is to utilize satellite imagery to detect areas that have crop lands by employing computer vision and machine learning techniques to analyze these images.


Q: What’s the status of your e-GUIDE projects thus far?
A: My project is still in its initial phases which currently consists of lots of literature review, reading through research papers, looking at a bunch of datasets and playing around with a few APIs


Q: What are your materials and methods for accomplishing your project?
A: That is still being determined but so far I will need to know locations of crop lands in Kenya therefore lots of relevant datasets from the Ministry of Agriculture are necessary. I will also require updated satellite imagery for these locations.

Q: What are your major challenges in carrying out these projects?
A: Getting the right datasets might be a serious challenge.


Q: Have you figured out how to overcome these challenges?
A: Not yet. I am still relatively new to research but from what I have learnt from my previous successfully completed research project, the more you read, the easier it becomes. However, you need to be selective in what exactly you read because there is so much out there


Q: What do you think the impact of your specific project will be?
A: That’s a tough one. I think my project right now does not seem related to electricity but rather identification of crop yields. However, this will make it easier to determine electricity methods most suitable for a given region given its agricultural crop focus. This is information helpful to planners to help them in selecting the right infrastructure for energy provision


Q: How do you think your project compliments other e-GUIDE projects?
A: My project opens up e-GUIDE to a different perspective in that it forces the initiative to ask itself questions that press beyond simply lighting up homesteads. My project goes beyond electricity for lighting and residential use but rather more productive uses in agriculture such as irrigation and processing.


Q: What do you like most about your project?
A: I love programming, building machine learning tools which I get to do a lot during the course of this research. It also gives me immense joy that my work is going to impact communities like the one I grew up in greatly transforming people’s lives.


Interview with PhD Student - Terry Conlon
By Joel Mugyenyi


Q: Brief background about yourself/PhD program?
A: I am in my fourth year of my PhD program at Columbia. I came here right after my undergrad at Duke where I studied mechanical engineering.
My research has been centered on two distinct tracks with the first track focused on working with US electric grids primarily in New York state parametrizing the grid and putting the grid into a bunch of high renewable energy penetration scenarios.
The other track concerns the electricity consumption in East Africa involving understanding residential electricity consumption using satellite images.


Q: Why did you decide to pursue a PhD/Research?
A: I always enjoyed the fact that in research you can largely design and pursue your own problems. There is a degree of freedom that you gain when first designing your project and the freedom you get from being immersed in some research topics. This is what primarily drove me towards research.


Q: What does e-GUIDE mean to you?
A: e-GUIDE is a very valuable community of researchers and experts who are looking to understand the drivers of electricity consumption in developing economies. I find this interesting because it is very open ended and there is a very large need for thinking and research into what potential future infrastructure should be put in place. I find that e-GUIDE is an excellent avenue for these challenges to be addressed.


Q: What do you think will be the most significant impact of e-GUIDE?
A: I think the most fundamental impact of e-GUIDE will be an understanding of the drivers of electricity consumption in the developing world as well an understanding of the best types of infrastructure to meet various energy systems goals such as electrification or reliability. I think e-GUIDE is uniquely suited to offer a range of expertise and computing tools to address these challenges that are currently not readily available to energy providers in the developing world.


Q: Tell me about your e-GUIDE project?
A: My research is first going to revolve around using satellite imagery to identify irrigation sites in Ethiopia using different types of remote sensed data that will involve various computer vision and deep learning approaches to process and understand that data. Plus an understanding of geospatial tools to comprehend and map the results of the analysis which we will feed into various planning studies in the future.


Q: What’s the status of your e-GUIDE projects thus far?
A: It’s in its beginning stages. It has been challenging obtaining ground truth data for irrigation sites in Ethiopia. I have been looking at the various forms of data available and the different APIs for different remote sensed projects. I along with the rest of the team that has been working on this currently have the techniques and methodologies that we are looking to use which we have figured out over the past few months.


Q: What are your materials and methods for accomplishing your project?
A: The materials will be the different remote sensed data with a combination of satellite imagery, radar data and various other products. The methodology will be geospatial analysis and processing along with various deep learning approaches to be able to figure out predictive methods to locate irrigation sites and later on grid planning and grid optimization methodologies as well.


Q: What are your major challenges in carrying out these projects?
A: Data management will be a challenge given that we are working with heaps of satellite images and remote sensed data which will be a massive data set. Therefore obtaining and storage of this data to allow easy access will be a challenge.


Q: Have you figured out how to overcome these challenges?
A: There is a lot of experience in the team especially in terms of cloud computing which I will draw upon and there are various resources online and at Columbia that will be useful in navigating relevant APIs. I also plan on taking a class on remote sensing tools that will be helpful.


Q: What do you think the impact of your specific project will be?
A: I hope to add to a general understanding of the different engineering problems that can be solved with remote sensing. I think there is definitely a gap between some of these more advanced tools such as processing of remote imagery and deep learning and the impact they could have on concrete engineering problems like grid expansion and increasing electricity reliability. I am hoping for a way to bridge concrete engineering problems and the more theoretical academic methodologies.


Q: How do you think your project compliments other e-GUIDE projects?
A: I think my project will hopefully provide a better understanding of agricultural and productive loads which will compliment the other projects like the work being done on residential consumption and grid planning.


Q: What do you like most about your project?
A: I love both the data and the methods that I am using . I think there is great potential in using satellite imagery time series data for understanding conditions on the ground. I also like these various deep learning methods that are able to put together highly accurate predictions from mountains for data.


Interview with PhD Student - Simone Fobi
By Joel Mugyenyi


Q: Brief background about yourself/PhD program?
A: I am a PhD student trying to graduate at Columbia University at the Mechanical Engineering department in the Quadracci Sustainable Engineering Lab (QSEL). My research is focused on understanding electricity consumption and prediction for developing countries. Prior to joining Columbia, I was at IBM in Kenya for a year with Prof Taneja working on energy systems and health care. Before that I was completing my Masters at Stanford which is where I met Prof. Taneja. Currently, I am in New Mexico for an internship with a company that uses satellite imagery data for different purposes.
I am originally from Cameroon but I have moved a round a bit spending some time in Burkina Faso, Ghana and Togo.


Q: Why did you decide to pursue a PhD/Research?
A: It was not evident to me initially that most of the work I was doing was tied to research. I find that I am a curious person who follows my interests. The ability to explore things that are not necessarily answered or fully defined has been interesting to me and over the years its become more evident to me that this aligns with a research path.


Q: What does e-GUIDE mean to you?
A: e-GUIDE is exciting to me because its an opportunity to leverage existing data sources to answer some the questions that developing countries will face in the coming years in terms of energy availability. I think in the past questions about generation and supply which were important were the main focus but I think e-GUIDE takes an alternative perspective of figuring out consumption patterns and how they change over time. e-GUIDE provides a space where we can holistically consider what the energy situation for developing countries can look like without necessarily following the path taken by more developed countries.


Q: What do you think will be the most significant impact of e-GUIDE?
A: I would like to see e-GUIDE working with utilities and influencing how they think about energy provision. There is an opportunity to help utilities answer questions that they might not necessarily have the technical expertise and resources to address.


Q: Tell me about your e-GUIDE project?
A: My project is about electricity demand prediction using satellite imagery. My hypothesis is that there are a lot of intuitions about electricity consumption and wealth which is linked to electricity consumption that can be obtained from satellite images. Factors such as road quality, roof top types, building quality can be indicators of wealth and serve as a good proxy for electricity consumption. Therefore my project is trying to answer the question of how far can we get with satellite images in terms of estimating electricity consumption and prediction.


Q: What’s the status of your e-GUIDE project thus far?
A: We have been able to answer some preliminary questions such as what is the link between satellite images and consumption levels. I would say looking forward my next task is to build a more robust predictive model and have a way to quantify the level of uncertainty within the model


Q: What are your materials and methods for accomplishing your project?
A: My most critical parameter is data from both the satellite imagery side and utility side. I also need to leverage huge clusters of computers to be able to run these computations.


Q: What are your major challenges in carrying out these projects?
A: My major challenges rest in the space of data. For example data quality really affects the model performance if there is huge corruption in the data I am working with. Another challenge is data availability especially when we try to combine ground datasets with satellite images where the images we currently have are outdated in frequency or resolution


Q: Have you figured out how to overcome these challenges?
A: I think of it as a bucket of solutions. For example I found that visualizing the data helps me pick out things that seem like anomalies. There are corrective techniques I can apply when combining satellite and ground data sets. I also read papers to find what sort of tricks have worked especially in the field of computer vision and machine learning with the goal of figuring out what can be useful to me.


Q: What do you think the impact of your specific project will be?
A: In an ideal case, what I would like as an output would be some type of portal where one could see different consumption predictions at different times for different developing nations. In terms of impact, I would like this research to be folded in by energy providers prior to building systems and over time it could possibly shift from an access platform to more of an operations platform since a lot of data goes into it allowing users to better optimize their systems.


Q: How do you think your project compliments other e-GUIDE projects?
A: I think the different projects are all trying to get to the same goal which is to try and understand what consumption looks like for developing countries at all consumption levels and how that affects an energy provider’s planning decisions.


Q: What do you like most about your project?
A: This is something that I have been quite keen about since 2014 when I first met Prof. Taneja. It is hard to pick anything about this project that I do not like. It’s difficult to pick one thing that is my favorite since I have been keen about it for such a long time.


Interview with PhD Student - Santiago Correa
By Joel Mugyenyi


Q: Brief background about yourself/PhD program?
A: I was working in Columbia with a utility developing a SCADA system for communications focusing on the generation perspective. I came to UMass with a goal of improving my computing skills which is necessary in setting up smart grids in developing countries.


Q: Why did you decide to pursue a PhD/Research?
A: I wanted to make an impact in my community. I am from a developing country and I had the chance to witness all the difficulties rural areas face in order to have access to electricity or a reliable power supply. I really wanted to study how I could improve the energy access for these communities.


Q: What does e-GUIDE mean to you?
A: It’s a very neat opportunity to understand what’s happening with the energy systems in emerging economies enabling a team of dedicated researchers to use this information to figure out how to improve grid systems using computing tools.


Q: What do you think will be the most significant impact of e-GUIDE?
A: I think getting the diagnosis of what’s happening with these grids and what has to be done to address the challenges being faced. This will have a significant impact on the grid and help communities that suffer from unreliable supply or very high costs of electricity


Q: Tell me about your e-GUIDE project?
A: I have been working on two main projects. In my first year, I was working mostly on reliability. I was not directly associated with e-GUIDE back then. I was developing statistical models and carrying out data exploration to measure power outages using smart phones with GridWatch.
My second project was focused on energy access using data from Solar Home Systems. In the analysis we discovered that there is a lot of energy curtailment with Solar Home Systems so the approach we took was to figure out how we could share this excess energy between neighbors.


Q: What’s the status of your e-GUIDE projects thus far?
A: Currently in the process of trying to publish a paper on sharing excess energy between neighbors using Solar Home Systems.

Q: What are your materials and methods for accomplishing your project?
A: I rely a lot on statistics because most of my projects are data driven. Stochastic process are also heavily used in my research. I am looking to use gang theory in my current research. In terms of methodologies, relying on heavier statistical algorithms and new trends such as reinforcement learning and deep learning


Q: What are your major challenges in carrying out these projects?
A: I think being novel in your research is tricky. It takes a lot of time and effort to find good research questions. Also being able to find the right path for a methodology can be challenging.


Q: Have you figured out how to overcome these challenges?
A: A very collaborative environment is important. It’s crucial to share your thoughts with your peers especially as you try figure out a methodology.


Q: What do you think the impact of your specific project will be?
A: Using data that helps us better understand energy systems in the developing world and address electricity access and reliability challenges I think will have a huge positive socio-economic impact on communities in the developing world.


Q: How do you think your project compliments other e-GUIDE projects?
A: I think taking advantage of the energy curtailment in communities using Solar Home Systems avails energy that can be used for a variety of new purposes from electro mobility to agri-processing.


Q: What do you like most about your project?
A: I really think the main motivation for doing my research is to be able to generate an impact on disadvantaged communities. On the technical part, I like to work on data driven projects where I can employ many computing tools such as AI, reinforcement learning.


Interview with PhD Student - Jorge Izar
By Joel Mugyenyi


Q: Brief background about yourself/PhD program?
A: I am originally from Mexico. I started my PhD program at Carnegie Mellon University three years ago. I did my undergrad in Mexico in chemical engineering and thereafter I worked for two years in Rio state. I then moved to the Netherlands to do a Masters degree in Sustainable technology because I really liked the idea of reusing waste. I pursued this based off an earlier idea of mine of designing a bio-reactor to produce bio-gas using manure.
After my Masters, I went back to Mexico where I started an initiative with family members where we produced organic fertilizers from red worms. I did this for four years before moving to the USA to start my PhD program.


Q: Why did you decide to pursue a PhD/Research?
A: Research is the starting point on the journey to improvement and changing pre-established common practices. I believe that only through research can we start to change and improve things pretty much through trial and error. It goes beyond academia for example my dad had a pest problem and through research he was able to identify solutions to address his particular challenge.


Q: What does e-GUIDE mean to you?
A: e-GUIDE is an initiative trying to better understand industry practices which in my case are agricultural practices and figuring out the most efficient methods that can be applied from both technical and non-technical perspectives so communities can realize an improvement in their social-economic standing.


Q: What do you think will be the most significant impact of e-GUIDE?
A: It think this initiative could be a success if in some years these studies are useful for people to better their living conditions where by decision makers can implement some insights from our work to make better use of power and agriculture systems.


Q: Tell me about your e-GUIDE project?
A: My side of the e-GUIDE project is related more to agriculture. I am currently focused on agricultural modeling working with Rwanda as a case study with a goal of understanding the agricultural systems in that country. We are using simulators to figure out ways in which certain crops yields can be improved.


Q: What’s the status of your e-GUIDE projects thus far?
A: We are preparing preliminary results to be presented to the Rockefeller foundation which will compare rainy and dry seasons for maize yields using historical data. The goal is to provide some interesting insights into the difference in yield realized when using and not using irrigation comparing both dry and rainy seasons plus a look into the power necessary to allow irrigation. Insights will include figuring out the major inhibitors to high crop yields such as insufficient water or fertilizers or perhaps the kind of cultivar used by farmers.


Q: What are your materials and methods for accomplishing your project?
A: One objective was to find out how much water was necessary to produce a certain yield so we implemented a variety of strategies to figure this out. In the beginning we planned on using a statistical model (regression based model) but it was not possible because we needed a high resolution model but there was not sufficient data to create a robust one.
Ultimately we found that crop growth simulators provided an excellent opportunity to address this challenge and we settled on one simulator from the FAO.


Q: What are your major challenges in carrying out these projects?
A: One of the main challenges was to get comfortable with new methods stemming from the lack of historical data for agriculture making it difficult to use statistical models. I thought about using remote sensing images and applying deep learning techniques but that proved not viable as well since we needed ground truth data. That left simulators as our most viable option which I was a little reluctant to use since they are quite complex but fortunately I found one that was not as difficult to manipulate. It is also still a challenge to get relevant data necessary to validate our models.


Q: Have you figured out how to overcome these challenges?
A: It’s still a work in progress. We figured out which method we wanted to use through lots of iterations which took about six months. Through collaboration with my advisors and members from the e-GUIDE team, we discovered different options that helped us address some of our challenges.


Q: What do you think the impact of your specific project will be?
A: In the short term, I hope that I can leave the next person after me a firm foundation to keep building upon and keep this important research moving forward.
In the long term, I would like this work to have a positive socio-economic impact on the people that live in these communities.


Q: How do you think your project compliments other e-GUIDE projects?
A: I think my project is a little different from the rest since I am more focused on Agriculture and the rest are a little bit more on the electricity side. I think my project provides a demand number for decisions makers on the electricity provision side on what exactly are the best energy infrastructure necessary to meet energy demand needs in the agricultural space.


Q: What do you like most about your project?
A: I really like the novelty of the project and learning new methods and ways of approaching different systems and challenges


Interview with PhD Student - Bob Muhwezi
By Joel Mugyenyi


Q: Brief background about yourself/PhD program?
I am from Rwanda. I previously worked with EDCL in Rwanda working on optimization models for long term capacity expansion of generation looking at the least cost combination of generation plants up to 2030. I also worked on transmission expansion projects for both HV and MV lines. I then joined STIMA labs at the start of 2019.


Q: Why did you decide to pursue a PhD/Research?
The work STIMA lab does is very close to the kind of work I was doing in Rwanda and I realized the research being done here could answer many of the challenges I was facing in my day to day work in Rwanda. For example, in demand prediction we were mainly relying on European models that were not well suited for the demand growth patterns in Rwanda. STIMA lab work is focused on the challenges unique to energy access in the developing world which is what drew me to them.


Q: What does e-GUIDE mean to you?
In my perspective, e-GUIDE is an initiative that addresses energy challenges in the developing world context ranging from electricity access, planning to effectively utilizing the limited resources available and ensure sustainability in energy provision.


Q: What do you think will be the most significant impact of e-GUIDE?
I think the demand prediction work will be helpful for utilities to come up with best practices.


Q: Tell me about your e-GUIDE project?
My work is not quite clear cut since I am at the beginning of my program. At the moment I am analyzing data for small commercial consumers of electricity looking at patterns in consumptions over the past couple of years. I am studying various factors that could affect their electricity consumption such as geospatial location, year of connection etc


Q: What’s the status of your e-GUIDE projects thus far?
I am working on data analysis looking for trends in a variety of consumption data.

Q: What are your materials and methods for accomplishing your project?
I will be employing python and python libraries to carry out data analysis for the initial stage of the project


Q: What are your major challenges in carrying out these projects?
Since I am at the start of my program, I will need to get comfortable with balancing my classes and research work in a way that I can offer each sufficient attention.


Q: Have you figured out how to overcome these challenges?
Not yet but I am in the process of figuring out how to best balance my classes and research work


Q: What do you think the impact of your specific project will be?
I think the best impact will be how utilities carry out demand prediction for communities that have never had electricity



Q: What do you like most about your project?
I get to work with lots of datasets. Having worked in electricity planning previously, I am excited about the impact my work will have on electricity utilities.