The Road to Radar

July 21, 2023
|
Palo Alto, CA

Array Labs is developing radar satellite clusters for the first real-time, high-quality 3D Earth model, enhancing affordable 3D data use in AR/XR, defense, climate, and insurance.

The journey from the clean-sheet idea to the launch of our satellites is a long one, and we’ve got plenty of hard work ahead of us. That said, we’ve come a long way in the first half of 2023, and we wanted to pause for a moment and briefly reflect on how things are going…

A snapshot of our progress

The story so far…

Before we can integrate our technology into prototypes and then production hardware, and eventually fly this satellite and its dozens of siblings in a cluster formation on orbit, we had to develop and demonstrate the core technology for our synthetic aperture radar (SAR) image formation system. 

To that end, Array’s first order of business is to create a synthetic aperture radar testbed to prove out the software and algorithms we’re planning to use on our orbital satellite constellation. A rendering of that testbed is shown below.

Array Labs Synthetic Aperture Testbed

This CAD file came to life over the course of a few weeks after we finalized our design and hired a supplier in Kansas. 

When fully articulated, the structure stands at 16 feet tall (not including the trailer). Here’s the system extending outwards for one of its first times while still out in Kansas:

With the rig built to spec and tested to our liking, we then prepared it for a long cross-country road trip to Array Labs HQ. To transport it, we folded the system; strapped everything down; and hit the road. 

Day 0 – The rig arrives.

On May 2nd, the test rig arrived in Silicon Valley. We took delivery in our parking lot. 

We took a brief second to marvel at the spectacle of the system, before moving it into our facility through the garage door.

This is where the fun begins. Up next on our checklist was preparing to convert this metal scaffolding into a space radar test rig.

Day 54 – First movement!

We achieved the first movement with the rig on Sunday, June 25th. As you can see, our workstation was not initially attached to our rail, requiring a human minder to enable movement across the room. Precise position and velocity control followed a few days later, with safety-critical sensors and automated test patterns close behind. 

Getting the lower stage online quickly allowed our RF team to get to testing our antennas and image formation algorithms. At the same time, we continued working on our upper stage.

Day 58 - First SAR data collected!

As we began collecting SAR data, which you can see above, we also were developing image formation algorithms to make sense of what our radar was seeing downrange. 

July 3rd: Working on image formation algorithms
The road to a successful SAR image…

Day 64 — We built a radar.

We developed our very own synthetic aperture radar, or SAR, on Wednesday, July 5th, after a breakthrough session of late-night image formation work. 

Left: radar reading / Right: overhead view of our test range

What you see above is the very first clean radar image from our indoor RF test range. The large red dot in the center is a corner reflector, located about 11 feet downrange from our radar. You can see the range in the image below.  

SAR hardware up close

When you put it all together, you get the following:

We’re currently moving from a single-channel radar system up to an eventual 32-channel system, which will allow us to move from 2D imagery up to a full 3D perspective. 

Founding engineering team shoutouts:

The above work would not be possible without incredible contributions from the entire Array Labs engineering team. Specific praise goes to the following members of the founding engineering team. 

The RF and hardware integration efforts are led, respectively, by Arun Nayagam and Max Perham, employees 1 and 2. 

Arun Nayagam

Arun joins us by way of the semiconductor and telecommunications industry, with 17+ years of experience across multiple companies, including established companies like Qualcomm and Amazon, as well as startups such as Intellon and Quantenna (both of which successfully IPO’d and were subsequently acquired). 

He has built and managed cross-functional teams involved in the architecture, design, post-silicon optimization, and productization of some exciting technologies like the first-ever Gigabit-class multi-input multi-output (MIMO) powerline modem, the first-ever 8x8 True-160MHz Wi-Fi SoC, and an industry-leading analog-compute based AI accelerator capable of running state-of-the-art deep neural networks at only 3-4 Watts. IP developed by him and his team can be found in over 100 million modems worldwide. Arun has numerous patents, issued and pending, in the areas of link adaptation and beamforming for powerline communications. He holds a Ph.D. from the University of Florida.

Max Perham

Max is a fellow YC alumni with a MS degree in Aerospace Engineering from Stanford University, who worked on Maxar’s latest and greatest earth observation satellite, Worldview Legion, before leaving to cofound Mezli in 2020.

At Mezli (YC W21), Max led the hardware team as they built and deployed the world’s first fully autonomous restaurant. The team created the entire 26 ft’ long fully robotic unit in under a year and watched as it served ~10,000 happy customers in San Francisco over a 4 month deployment in the fall of 2022. Throughout its deployment at the Spark Social food park it boasted a greater than 99 percent uptime, and did not shut down at all for over four straight weeks, a pair of stats which Max is especially proud of.

The Mezli autonomous restaurant: 99%+ uptime over 4 months

Divyanshu Jain (DJ)

DJ is our lead FPGA engineer and the fourth employee at Array Labs. 

After graduating with a Masters in Electrical and Computer Engineering from UCLA, DJ joined Mojix, a startup making RFID readers. DJ was involved in all aspects of design and implementation of communication and DSP algorithms on FPGA and MCUs during his six years at the company. He then joined Broadcom as a system design engineer and was involved in developing High-Level Synthesis (HLS) infrastructure for chip design. He was also involved in the design of ultra low power ICs targeting Bluetooth, Zigbee and NFC technologies. DJ later joined NXP and worked on the digital design for NXP's first Ultra Wide Band IC targeting secure localization. He was also involved in developing RADAR applications for UWB. DJ spent 10 years designing wireless communication ICs that are used by all major cellphone vendors. DJ is skilled in designing, implementing  and testing ultra low power ASICs that meet power, performance and area benchmarks.

Shadi Youssef

Shadi is our resident RF expert and employee # 5 at Array Labs. 

Shadi's expertise spans more than 13 years in the field of RF and analog design. From 2006 to 2009, Shadi was with SiTel Semiconductor in the Netherlands (currently Dialog Semiconductor) working on RF CMOS power amplifier design and integration. After receiving his PhD from the University of Twente in the Netherlands in 2013, Shadi joined Marvell Semiconductor in Santa Clara where he worked on the development of their flagship WiFi chipsets. From 2017 to 2021, Shadi was with Mojo Vision in Saratoga, a startup with the ambitious goal of building a smart contact lens for augmented reality applications. During his time at Mojo, he led the radio chip development efforts from architecture inception all the way to chip design and validation. The successful design was a key part of the feature complete contact lens demo that the company unveiled in 2022. In 2021, Shadi joined Meta's Reality Labs where he led the WiFi development efforts within the RF Technology Group for forward looking augmented reality projects.