Where Data and People Meet

An illustration showing multiple semi-transparent screens with data against a navy background with columns of faint zeroes and ones, binary code.

A team explains the highly collaborative Roman Research Nexus, including its future impact on science.

About This Article

For the astronomy community, the launch of the Nancy Grace Roman Space Telescope is right around the corner — no later than May 2027. The telescope’s large surveys of the sky will produce an unprecedented amount of data, which has been a driving force for the institute’s engineering division as staff continue to help develop and mature the Roman Research Nexus, the place where astronomers will access and analyze its data.

In 2024, approximately 30 engineers and scientists at the institute collaborated to build out a “meeting spot” for future Roman telescope data and its users. Three members of the team, astronomer Claire Murray and cloud engineers Prem Mishra and Manuel Sanchez, reflect on the work achieved this year and the impact the platform will have on the mission, the science community, and how research is done.

Why is it essential that Roman has a science platform? How does it benefit researchers?

Claire Murray: The Roman mission is unique compared to the other missions that the institute supports, because it is a survey mission. Roman has an extremely wide field of view and a fast survey speed. Those two facts mean the data volume is going to be orders of magnitude larger than what we're used to, which impacts how we host and distribute the data to researchers. We will enable users to interact with this gigantic dataset, which is estimated to be on the order of 30 petabytes, on a science platform in the cloud. Rather than users downloading the data, we bring the data to the users. They will be able to log in to the platform and perform the same types of analysis they would normally do on their local machines.

Prem Mishra: Until now, astronomy has typically required researchers to pull down the data. With Roman it won't be possible to do that. It's critical to keep the computing power and data together so you can do high-volume, high-speed data processing. And that's what the science platform achieves. They stay together so innovative work can be done at speed and scale, and produce results much faster.

Manuel Sanchez: Working in the cloud also means users don't need a powerful computer. Think of users in remote areas or who lack grant funding. All they need is a good internet connection and an email to do research. 

Share with us some of the major accomplishments achieved this year, along with ongoing challenges.

Murray: Users can now organize into groups and share resources on the platform. This differs from the previous setup, where users were in their own individual ecosystems with their own home directory, files, and software. Now a group can share storage, computing resources, and software, no matter where they are in the world.

Mishra: We want this to be as simple as possible for users, which means the backend engineering is strong. To enable collaboration, we built an authentication system that “talks to” our grants management and user management systems behind the scenes. This lowers the barrier of entry. Since we want to allow users to log in only with their email, which increases the security risk, we also had to reinforce our infrastructure so it’s safe.

Sanchez: To add to what Prem said, a user logs in and it simply works. It looks as though nothing really happened. But to make it look simple in the front end, there are a lot of moving parts in the backend. That goes for the system’s scalability as well. There are many different parts involved in the scalability, it's not just processing power and storage. Everything has to work so it appears uncomplicated to the user.

The institute supports other missions’ science platforms. How did your team leverage those experiences to shape Roman’s platform?

Sanchez: The Transiting Exoplanet Survey Satellite (TESS) mission’s science platform, known as the Timeseries Integrated Knowledge Engine (TIKE), was our first try at introducing users to this type of workflow on the cloud. With TIKE, we have been learning researchers’ usage patterns. We can track the performance and metrics, which is helping us design the appropriate environment and capabilities for Roman. Our experience with other science platforms also helps us save time from a coding perspective. The code is basically the same for our platforms, but we can customize it as needed.

How does the science community play a role in the platform?

Murray: Recently we've conducted workshops so people can work in the platform. Our first workshop was in June. Folks from different areas of the institute, such as the Roman Telescope Branch, the Instruments Division, and the Data Management Division, tried out the platform. This workshop gave us our first round of feedback on both the infrastructure and content.

In September, we hosted another workshop for members of the proposal infrastructure and Wide Field Instrument science teams, as well as the Project Science Office at NASA Goddard Space Flight Center. Over two days, they tested the platform and tutorial content. Since these are Roman experts, they really had great ideas of what they wanted to do.

We were really excited about the January 2025 AAS meeting, where we had another workshop that was open to everyone in the astronomical community. This was our first opportunity to get new eyes on the platform, particularly those who aren't affiliated with the institute or official Roman teams. Overall, we’re really looking forward to getting a better sense of how an average user will interact with the platform. Accessing this cloud-based platform is going to be a big leap for most people, so I think we're going to learn a lot.

With the launch of Roman rapidly approaching, how are you feeling?

Murray: I'm personally excited to be part of a mission that is so collaborative and focused on open science. We still have a lot to do. We're making all of these infrastructure upgrades with a user support-focused mindset. We have to think about external users more than we ever have before. Looking ahead, it's going to be a busy time.

Mishra: I’m excited and a bit nervous as well. Roman is the biggest and only mission that will be doing everything in the cloud. That is monumental. Our focus remains on making sure the Science Operations Center is ready for this, and that we are ready to support the astronomy community. That's a huge responsibility.

It should go without saying, but this work is not easy. From the engineering aspect, it takes a lot of effort, patience, and coordination. Teamwork is what makes our progress possible. I know there will be challenges, but we have a strong team. I'm also thankful that the institute is able to do this because Roman’s data is going to change how astronomy is done.
 

Against a wavy, navy blue background with a scattering of small white dots are illustrations of the Hubble and Nancy Grace Roman space telescopes, along with blue-green cubes representing the amount of data they return per day. Toward the top left is Hubble, which has a gray cylindrical body and two blue rectangular panels. Four columns of faint zeroes and ones, binary code, stream into a single blue-green cube toward the telescope’s lower right. The cube is labeled “2.7 gigabytes per day.” Toward the top right is Roman, which has a gray barrel-shaped body, orange sun shade, and blue solar array shield. Many columns of binary code stream down into a large tiered stack of 452 blue-green cubes. It is labeled “1,375 gigabytes per day.”
With its extremely wide field of view and fast survey speed, the Roman Space Telescope will produce an unprecedented amount of data, impacting how we host and distribute the data to researchers. While the Hubble Space Telescope produces 2.7 gigabytes per day, the expected daily data rate for Roman is 1,375 gigabytes. Scientists estimate Roman will return 30 petabytes of data in its first five years of operations.

Share This Page