Dan Berkenstock: The world is one big dataset. Now, how to photograph it ... | Summary and Q&A

TL;DR

This content discusses the journey of the author from being a Ph.D. student to building a new generation of imaging satellites for global transparency.

Key Insights

• 🛰️ Satellite imagery has been underutilized in data analysis, despite the vast amount of sensors surrounding us today.
• 🤔 The majority of satellite imagery available is outdated, which limits its relevance to current work and research.
• 🌍 Only a small fraction of Earth's surface is regularly captured by satellite imagery, leaving large areas unseen or barely seen.
• ✨ The cost of traditional satellite launches and the desire for functionality have led to larger and more expensive satellites.
• 🚀 The concept of hitchhiking small satellites alongside larger ones has significantly reduced the cost of space exploration.
• 🔬 The laws of physics dictate that a larger telescope is necessary for better picture quality in satellite imagery.
• 💡 Combining individually noisier frames into a videostream allows for high-quality satellite images, at a fraction of traditional costs.
• 📡 Launching a constellation of satellites will allow for the collection of petabytes of data and the development of scalable analytics for insights.

Transcript

Read and summarize the transcript of this video on Glasp Reader (beta).

Questions & Answers

Q: What inspired the speaker to start using satellite imagery in their work as a data scientist?

The speaker was inspired to start using satellite imagery in their work as a data scientist after realizing that all data has a location and that satellite imagery could provide valuable information. They were intrigued by the potential to answer questions about the global economy, such as the amount of oil stored in China or the number of ships in world ports, using up-to-date imagery.

Q: Why were the pictures obtained from satellite imagery years out of date?

The pictures obtained from satellite imagery were years out of date because the satellites used for imagery were expensive to launch and there were not many of them. The high cost of launching satellites led to infrequent launches and a focus on cramming as much functionality as possible into each satellite. As a result, the satellites became larger and more expensive, which limited the availability of fresh satellite imagery.

Q: How did the speaker and their friends address the challenge of expensive satellite launches?

The speaker and their friends addressed the challenge of expensive satellite launches by exploring the concept of hitchhiking small satellites alongside larger satellites. This approach significantly reduced the cost of putting objects into space, allowing for experimentation and innovation. They designed their own satellite, taking inspiration from the smaller CubeSats, but also finding a compromise to ensure the satellite could capture high-quality images by building it slightly larger than a breadbox.

Q: How did the speaker overcome the limitations of small satellite telescopes?

The speaker and their team overcame the limitations of small satellite telescopes by using a minimum viable size of one meter for their telescope, which allowed them to capture higher quality images. They also utilized a new generation of video sensors, originally developed for night vision goggles, which allowed them to take video streams of noisy frames and recombine them into high-quality images using pixel processing techniques on the ground. This technique reduced complexity and cost compared to traditional line scanner systems.

Q: What is the purpose of building and launching these satellites?

The purpose of building and launching these satellites is to provide global transparency and unlock insights through timely and up-to-date satellite imagery. The satellites aim to gather petabytes of data, allowing for the analysis of economic data and the discovery of connections previously unknown. The speaker sees themselves as pioneers in this new frontier and believes that providing global transparency through satellite imagery is an idea whose time has come.

Summary & Key Takeaways

• The speaker used to work as a data scientist and satellite engineer, but did not consider using satellite imagery in their work until they realized the potential value of more recent pictures.

• The majority of satellite imagery available is outdated, and the high cost of launching and operating satellites limits the quantity and frequency of new imagery.

• The speaker and their team developed a new design for smaller, more affordable satellites called CubeSats, which can capture higher quality images through the use of video sensors and advanced pixel processing techniques.