kartikkumar

Posting this here because it's a fundamental feature missing in HipChat and the issue was created over a year ago. It's quite fascinating to watch how little is being done to address this and how many people have or are threatening to leave HipChat as a result. We ultimately made the decision to ditch HipChat as a result of this.

It's an initiative that has been backed by the Gates Foundation and Omidyar Network. There have been a few press releases about the goals of Radiant, e.g. [1][2]. I shared the link with the intent of simply pointing to their mission, which is to provide free access to a wealth of geospatial data.

I don't know where the project stands at present, so I dont have an objective measure to discriminate between what they claim to be doing and what they're actually doing.

Coincidentally, I had a meeting in DC today at the Omidyar Network and Radiant was mentioned briefly in the context of providing startups with a way to bootstrap without spending large chunks of money on acquiring geospatial data.

[1] http://www.reuters.com/article/us-technology-satellites-huma...

[2] https://www.independent.co.uk/news/science/bill-gates-melind...

Check out Radiant: https://www.radiant.earth

Fantastic achievement and a testament to the great work done by an army of engineers and scientists. 37 successful launches and counting; worth standing still for a moment and appreciating that level of success in an inherently risky business. Well done to ISRO and the PSLV team!

This. As great as it is to hear about sensational founders that have turned the world upside down, it is at times hard to relate to. It would be great to show the ground truth of what small companies have to go through, and having office-hours that help elucidate the thought processes to follow to make steps forward would be invaluable. The best podcast episodes I've heard (and I specifically say episodes because I don't think any show that I've heard has consistently achieved it) on this subject have just been really good at being down-to-Earth and relatable.

I've developed a lot of open-source space simulation code and got a few projects in the pipeline. Maybe worth having a chat? My email is in my profile. Anyone interested, feel free to ping me.

The radar was only activated after the front shield was ejected, so there was no altitude data prior to that event. It could be that accelerometers on board sensed the dynamic pressure, so that parachute deployment was not purely open-loop. I'll try to dig through publications to find out what the EDL (Entry-Descent-Landing) strategy was.

EDIT: Here are some papers worth persuing:

[1] https://solarsystem.nasa.gov/docs/Bayle_ExoMars_EDM_Overview...

[2] http://ippw2016.jhuapl.edu/docs/abstracts/Missions-abstracts...

[3] http://arc.aiaa.org/doi/abs/10.2514/6.2011-6341 (behind journal paywall)

It's not that it hadn't been hypothesized before, it's rather that the numbers that they attribute to this hypothesis stack up well against a lot of Solar System observations. In other words, the hypothesis is not merely that there is another large, undiscovered planet in the far reaches of the Solar System, but rather the details of its mass and orbit that make this a compelling case. Part of the reason that this hypothesis wasn't possible before is because the observational data to support these numbers was lacking. The amount of knowledge we've gained about the Kuiper Belt over the last decade or so is phenomenal and drives our understanding of what might have happened in the early Solar System. It's worth reading the original paper to get a handle on the rigour with which this hypothesis has been analyzed and the coupling with observations of the Kuiper Belt [1].

[1] https://arxiv.org/pdf/1304.5166.pdf

I use SQLite to store all of my simulation data (~10s of GBs). It's remarkably versatile and the fact that there are good libraries for Python and C++ to interface with and query SQLite dbs makes it a synch to use for data analysis.

I've seen so many people struggle with custom binary formats; I imagine there are countless research hours lost in figuring out how to work with these obscure formats. I've advocated to all students I work with to make use of SQLite to store simulation data for their thesis projects and my experience is that they're quick to pick it up and figure out how to do some pretty complex querying.

It's one of those things that I don't understand about academia: there are so many standards and well-established tools in the tech/IT sector that we don't take advantage of. SQLite and JSON are the two that I constantly advocate to everyone I work with.

Great links

I founded an open-source library for orbital mechanics during my PhD, which is still going strong: https://github.com/tudat

We're developing a new set of modular libraries based off of Tudat: https://github.com/openastro. We're also building an open search engine for satellite subsystems: https://satsearch.co.

Some other interesting open-source stuff going on:

- https://librespacefoundation.org

- http://www.kubos.co

- https://github.com/esa/pykep

- https://gmat.gsfc.nasa.gov

> last I heard they were growing lettuce in space...

And that is not interesting ... why? Growing food in space is an incredibly fascinating area of research and vital for human spaceflight beyond cis-lunar space. Additionally, micro-gravity offers a unique opportunity to study the core mechanisms for growth, with the necessary spin-off of helping develop more robust crops back on Earth.

Lots of open-source initiatives going on in the (small) satellite world at the moment. We're hopefully going to be releasing an online platform for preliminary design and trade-off analysis in the next few months. I know of a few other companies that are working on open-sourcing different parts of the technology stack. Predicting New Space to look quite different in another 5 years, with hopefully next to no barriers to entry.

Absolutely agree. Eerie particularly because of the lack of mention of ethical considerations in the same breath.

Yea a proof of concept would definitely help. We're trying to architect the system at a high level though for the purposes of this project, which is why I thought I'd ask around here for any insights.

Thanks!

That's really cool. Do you have anything published online at the moment?

Interesting idea! I've submitted the form. Might be easier to get a few more people to sign up if you provide a few more details about the problem you're addressing and what you're planning to build.

Yea it does seem like the potential for a massive leak. I wonder if employees/academics are even allowed to consent to things like this one university premises.

I would hope that "consent \neq legal". That opens up a huge can of worms, since they imply there would be harvesting EVERY url visted, not just hits on scopus, science direct etc.

I just received this email and can't believe it.

> In this study an Online Usage Tracker is applied to track which websites you visit, by recording the URLs.

How does this not violate EU privacy laws?

Thanks for the feedback! It's a slow process, but we're plugging away at trying to on-board users and pinpoint where our open stack can directly address challenges they're facing. Our best starting point at the moment seems to be universities building satellites, as we have a good network and the barriers to get started on evaluating product/market fit are lower than B2B.

We are indeed talking about the burgeoning sector that is often referred to as NewSpace, which is made up of universities, research institutes, agencies and commercial companies. The idea is to use our network in the Cubesat/PocketQube community to prove the benefits of integrating the spacecraft design phases, ultimately leading to more robust designs and cost savings. It's also a sector that is highly fragmented at the moment, with parts suppliers popping up all over the world.

If we reach the target of proving the benefits of our open stack, I think large industry players will have to take note. We've had some discussions with the likes of Airbus DS in Europe and the sales cycle is very long, but not impossible.

The mission that we're on is really to open up the sector to completely new applications and teams. We want to expose the tools and technologies that are needed, so that it's not just a select few that can reap the benefits of space technology and infrastructure.

I definitely find myself having to wade through my photo and video collection just to find that one that I remember I took a few years ago but for the life of me can't remember exactly when and where (which is the easiest way to sort and find). I think the key for me would be how much training/how many questions it would take for the search to become accurate. It would also be important to me to be able to use this offline. Don't fancy having to upload all my personal photos and videos.

Have you done any user testing yet?

I can imagine that if you get the algorithm to work robustly that there might actually be an entirely different, enterprise market.

Do you have a proprietary algorithm? Or are you using something well-known, published/publically available?

Would this simply be a product that the end user pays for? Or do you have another revenue stream in mind?

Thanks for the feedback.

Satellites are complex machines and there's a lot of engineering work that goes into figuring out exactly what to build. The goals for a mission are usually set by a client/end user: e.g., I want images of San Francisco at a resolution of 5m/pixel or less.

Engineering work goes towards simulation and analysis to figure out what the size of the solar panels are that you need to generate enough power on-board to power the camera, what the size of the battery needs to be to make sure that the satellite keeps functioning during eclipse, what the size of the thruster is that you need to keep the camera pointing at the right spot, etc.

Once you've figured out what you need, the job of finding a vendor that has your part involves wading through their websites to find data sheets, calling them, emailing them etc. It's a really time-consuming process and all you want to know is if there's a part that matches what you need, within your budget, that has heritage in space, etc.

The engineering work is iterative because the subsystems within the satellite all depend on each other. Hence, you're doing this search over and over again. Usually, you just end up making sub-optimal choices.

We've tried to explain the search problem in the following video: https://www.youtube.com/watch?v=FlHEoq7utW4. The low-hanging fruit comes from the fact that a lot of data is already available online, just completely unstructured.

The idea of tying this in with simulation tools is that it allows you to do things like programmatically figure out what happens to your design if the vendor you've picked for e.g., your battery fails to deliver. This kind of sensitivity and robustness analysis is very important during the design process, because it allows you to reduce uncertainty and figure out the main risk items that you need to devote engineering time to.

Bit a long explanation, but hopefully that gives you a better sense of the immediate problem we're trying to tackle for engineers.

For vendors, the pain simply comes from not knowing where their users are, i.e., teams that need the parts that they are selling.

Does this help explain the pain better?

A few more details due to the character limit:

BACKGROUND

There are a slew of new applications that can benefit from space technology, e.g., "internet everywhere", IoT (M2M messaging), asset tracking etc. Rapid growth of the small-sat sector and robust consumer electronics have precipitated a drastic drop in the cost of space-based infrastructure.

BUSINESS MODEL

We'd love to put together something like the Red Hat business model: delivering enterprise-level, open-source software under SLAs. With satsearch specifically, we plan on delivering a comprehensive analytics report (likely monthly) that allows suppliers to monitor what's being search for and how the market is evolving (under a subscription model).

We've already had a number of inbound requests from suppliers to be listed in satsearch. We are in advanced negotiations with a major space agency (direct procurement), a supplier that wants our structured data for market analysis, a university small-satellite team and two governmental organizations seeking to raise the profile of their space clusters.

Besides directly serving the space industry, we have supported an indie game developer in adopting our libraries for a new space-based strategy game. We have also had discussions with a company interested in leveraging our tools for tech transfer.

A lot of applications are much closer to home. For instance, I've seen a few teams pitch the idea of building a robust, small-satellite network to serve as a platform for Machine-2-Machine messaging. I think enabling the global IoT industry is really going to unleash some spectacular applications. Internet everywhere is of course a popular idea that's being worked on. I particularly like the low-data rate, low monthly charge model that I've seen pitched. Seen teams pitch internet anywhere for as low as a couple of bucks a month.

Being in the midst of all of this, I can vouch for the palpable excitement within the industry: at least if you stay away from talking to traditional prime contractors. There's a real sense that disruption is right around the corner. The biggest problem is that there have been no significant exits to prove investors right. This makes it difficult to raise capital.

Also, traditionally, external capital has sought out downstream opportunities. We've been focussing on upstream tech and it's much harder to prove the case given the long time horizon.

I've seen some incredible applications being pitched, but I think the real test will be whether they are able to generate significant exits. I, for one, hope that the next five years spells a period of immense productivity in the sector, generating profitability all along the value chain. If we're able to ride that wave, I think that there are many more applications that will pop up that make use of space infrastructure.

The blog post I linked to in my previous comment suggests differently. Do you happen to know where you saw that diagram? Since I'm working on space debris, I'm curious what the target graveyard orbit is if it's not escaping.

The upper stage gets left behind in Earth orbit. The Briz-M stage provides the kick to propel the payload to a higher orbit and then usually there's a kick motor that follows up, in this case, delivering a Trans-Mars Injection maneuver to set the spacecraft on its way to Mars.

EDIT: Actually realised I'm wrong since the Briz-M is the 4th stage, not the 3rd. The 3rd stage remains in orbit and then splashes down and the 4th stage actually provides the entire kick for TMI [1]

[1] http://blogs.esa.int/rocketscience/2016/03/13/why-exomars-ri...

I wonder if there's enough data to reconstruct the demise of the upper stage. Re-entry physics is extremely interesting and complex. ESA has a number of articles on the analysis of the re-entry of the GOCE satellite [1]. There are a couple of tools that are considered the defacto standards for computing high-altitude re-entry physics like SCARAB and ORSAT [2].

A colleague of mine working in the same EU project as me [3] is developing an open-source tool that can switch between low- and high-fidelity computations [4].

Wonder if any of the observations recorded are valuable to calibrate these models.

[1] http://blogs.esa.int/rocketscience/tag/re-entry/

[2] http://ntrs.nasa.gov/search.jsp?R=20100005304

[3] http://www.stardust2013.eu

[4] http://www.congrexprojects.com/Custom/15A01/Presentations/Au...

Ok great, thanks for clarifying!