Julius

I am a research scientist in Meta's Physical Modeling team. I work on technologies for data center sustainability.  Data Centers are very important, as they enable our own apps and services such as Facebook, Instagram, WhatsApp, Quest; Data Centers also run the internet, which are an important element of our lives as the internet connects people and ideas. Data Center construction has been growing rapidly in the past few years and will continue to do so into the future. Data Centers are existential to Meta, and sustainability is a key goal of our company. 

One example of my work is showcased on the Tech@Meta website, where I worked with research and industry experts to use AI to optimize concrete formulae. In collaboration with researchers at the University of Illinois at Urbana-Champaign, we have developed a new AI model that optimizes concrete mixtures for sustainability as well as strength. In early field testing, carbon emission was reduced by 40 percent, while strength requirements were exceeded. Cement in concrete accounts for approximately 8 percent of carbon emissions globally. If successful, the impact of this work could reach well beyond data center construction, as it applies more broadly to the general construction industry.

I first came to Facebook in 2018 to join the Facebook Connectivity Lab. I worked on technologies to analyze rural connectivity challenges, optimize wireless systems, and improve network deployments. You can read about some of my work here and here. In collaboration with colleagues and partners at Facebook and beyond, we have developed and published several works, including how to improve rural connectivity coverage using diffraction, use-case explorations using mmWave technology, wireless network modeling tools, open-sourced microwave link design software, and an analysis of the synergy between 5G and Wi-Fi 6. 

Prior to Facebook, I was a principal scientist and program manager at Schlumberger Research in Cambridge, MA, USA, where my specialization was on communication and sensing technologies for Subsurface and Subsea systems in the oil and gas industry. I worked on digital communication systems for drilling, wireline, and reservoir characterization, using electro-magnetics, acoustics, and wireline. My team and I were the first to demonstrate real-time wireless video at 1km over a subsea channel, using acoustic telemetry. Technologies that I developed are used for real-time monitoring and control of drilling and testing systems; data compression for geophysical data; signal processing and networking in the sub-surface. 

I received my doctorate from MIT in 2006, where I was a MIT Presidential Fellow. I also attended UC Berkeley where I was awarded the Demetri Angelakos award, and Purdue University.

Facebook Connectivity 

Facebook Connectivity's mission is to bring more people online to a faster internet, at the heart of Facebook's mission of giving people the power to build community and bring the world closer together. 

As The Inclusive Internet Index shows, the digital divide appears to be widening. Access to the Internet not only allows people to connect with each other. It also gives access to a wealth of information and the potential to innovate. 

Solving the problems of Connectivity requires that we understand how people live, how they can access the Internet, and how they want to use the Internet. Therefore, I am building a coalition of partners, ranging from academics, to researchers, to industry players, to domain experts. I have active projects in economics, analytics, numerical simulation, optimization, and wireless networks. 

One project that I completed was on using diffraction to design and deploy wireless backhaul networks. In rural areas, the cost of deploying fiber optics is prohibitive, therefore wireless backhaul is highly desirable. Working with partners at Internet para Todos and Mayu Telecommunicaciones in Peru, we deployed dozens of links using this technology, providing coverage to many settlements. 

In addition to network design, I worked on smart power management of off-grid sites. Power is a major barrier in telecom, and smart power management can reduce operational and as-built cost of telecom sites. This gives major sustainability wins in operational and embodied carbon. Field trial shows that 40% wins are possible. 

Schlumberger Research

Schlumberger is the leading oilfield technology & services company. It invests more in R&D than its two biggest competitors combined. I joined Schlumberger after my PhD because I wanted to have the opportunity to drive impact through technology, and I was fortunate to get exactly what I wanted. Schlumberger develops technologies that it then uses itself, to provide services to the oilfield: it takes an integrative approach to technology development. I learned that it's not only about building the fastest, biggest, shiniest technology, it's all about whether it solves real problems and provides real value. 

I started in Houston, however after two years I asked to move to the Boston area. Schlumberger started a new research group at the Schlumberger-Doll Research Center with me as the founding member. In the Telemetry Program, I developed Connectivity and Sensing technologies for Drilling, Wireline, Testing, and the Subsea. I have been granted over 11 patents from my work at Schlumberger, and 3 trade secrets. 

My team and I developed what we called Physics-Based Telemetry. Oilfield operations require Connectivity through very challenging communication channels, such as electro-magnetic signaling through the underground, acoustics in the subsea, and acoustics on metal conduits. We re-examined the basic physics of these channels, and use them to inspire algorithms for communication and compression. We developed receiver algorithms, noise cancellation methods, and compression algorithms that are inspired by physics. 

My team and I were the first to demonstrate real-time wireless video at 1km over a subsea channel, using acoustic telemetry. We were able to establish this world record within 18 months, thanks to excellent collaborations with Lee Freitag and his team at the Woods Hole Oceanographic Institute, and with Jim Preisig. 

MIT, UC Berkeley and Personal Research 

My PhD thesis at MIT was on signal representation and acquisition. 

My MS thesis at UC Berkeley was on distributed compression and multi-user coding. 

I have been fortunate to have the opportunity to work with Lav Varshney and Ram Srinivasan, whom I met at MIT. Lav beautifully extended a counter-example provided by Bob Gallager on representation, compression, and malleability of information. This turns out to give insights into the fundamental relationship (and conflict) between compression and updates of data; it further developed into a theory on cloud storage. Ram discovered that my work on signal acquisition has relevance to neural signal processing, and built a brilliant body of work exploring its applications.