PAY Meeting 2026-02-28

<aside>

This page seeks to describe a vision for coordinating efforts between subsystems within the PAYLOAD group of UTAT Space Systems, which include Science, Optics, Optomechanics, Pay-Elec, and Data Processing. This group is (currently) developing the FINCH EYE instrument, which is a hyperspectral Earth observation SWIR imager for crop residue abundance estimation.

</aside>

Motivation and Purpose:

A cognizant division of Space Systems into the BUS and payload (PAY) groups will enabled more focused progress towards a functioning satellite. The BUS and PAY groups can “handshake” through a robust volume, mass, power, and data budget, as well as the appropriate mechanical and electrical interfaces.

The BUS group will seek to develop a more general-purpose satellite BUS that can be re-used and improved upon over the years. They will also allocate volume, mass, and power for a potential payload. The PAY group will create a payload to fit within these specifications, with some room for negotiation and fine-tuning.

The current goal of the PAY group is to build FINCH EYE and integrate it within the satellite mainframe provided by the BUS team. However, if things “don’t go well” with the payload, the BUS team can continue without - they can find someone else who has / is ready to build a suitable payload or just downgrade from a 3U to a 1.5U satellite. With the technical risks disentangled, the BUS and PAY groups can move at their own pace and deliver a higher quality end result.

Our Main Goal in PAY…

…is to get the CubeSats Initiative in Canada for STEM (CUBICS) grant, with the deadline approaching in September 2026. The CUBICS grant is our best shot at getting the funding necessary to build and test the FINCH EYE payload, given that it’s aimed specifically at increasing university students’ participation in satellite design. Funds raised through UofT student levies are sufficient to “keep the lights on” and build prototypes, and should hence be used for such purposes only.

To strengthen our application for CUBICS, we need to secure “partners” or co-applicants for the grant. For example:

Dunlap Institute for Astronomy and Astrophysics: has staff optical and optomechanical engineers with experience building space-borne optical instruments. Their expertise would be critical in guiding the design and fabrication of the FINCH EYE hyperspectral payload. Although they may be less incentivized to contribute their time to a project tangentially related to their main line of work, we can entice them by offering an opportunity to build a payload for a future UTAT satellite mission.
Earth Sciences Dept. at UofT: a professor here could supervise research efforts within the Science team. Currently, we have a tenuous scientific value proposition, since most crop residue studies use 2-2.2µm and popular vegetation indices (e.g., NDVI) use the VNIR range. After completing research on crop residue abundance estimation, it may be helpful to examine other scientific use-cases for a hyperspectral imager in the SWIR (or VNIR) range.
Optical fabrication companies: who will manufacture the custom optics and optomechanics, as required for our project. They will supply us the quote, so we know how much it’ll cost to build the payload, and then ultimately deliver the product.
MDA: has expertise in assembly, integration, and testing, including optical “vision” systems. They would guide us through the shake & bake testing aspect, and provide high-level advice.
The Canadian Space Agency (CSA) itself would provide some technical oversight and support throughout the project, which would supplement contributions from our partners.

Without these partnerships, as described, our project may be dismissed as “too ambitious”. Our application to the FAST grant failed to make the cut because we didn’t have experts in optical design and machine learning. These partnerships will convince the grant committee we have adequate technical expertise to support our ambitions.

Our team would benefit greatly from tight collaboration with these partners. Realistically, it is good to have “adults in the room” when working on difficult engineering. Having decades of experience under their belt, we can rely on them to spot mistakes we are ignorant to and guide us along the right path.

That being said, it is also crucial these partners stick with the project until its completion. For example, it would be bad if optomechanical engineering experts at Dunlap decide to quit the project mid-way, right when we are approaching the fabrication stage of the mechanical housing. The burden falls on us, UTAT, to manage relationships with our partners, to keep them engaged and avoid misusing their time.

From March to May 2026, we will be actively reaching out, securing these partnerships, and negotiating the terms of the partnership should the grant be awarded to UTAT. Then, from June to August 2026, we will be working on the grant itself. We expect to know the decision sometime by June 2027. In the meantime, we would be working on prototypes…

Prototyping: A Roadmap of Minimum Viable Products

Perhaps the most exciting question to an engineering audience is “what are we going to build?”. Instead of building the payload “all at once”, it would be pragmatic to first build prototypes that let us practice and lean the ins and outs of the problem before committing large money towards final space hardware. We would be demonstrating to the CSA that we are capable of building a bicycle, so they can confidently invest in us to build a car.

| SWIR Spectrometer using FLIR TAU We can build a SWIR spectrometer made from COTS optics, the FLIR TAU, and a machined optomechanical housing. This lets us practice making a reflective spectral engine and become intimately acquainted with the challenges this design has. • Optics will develop the optical design and perform tolerance analysis; • Optomechanical will design and fabricate the housing, and integrate the optics together with Optics; • Firmware will develop a PCB that reads out data from the FLIR TAU and sends it to a laptop via USB; • Data Processing will create a GUI that displays the data collected with some processing options; • Science will use this spectrometer to collect ground truth measurements and study how moisture content & viewing angle affect the spectrum of soil and crop residue.

The sensor is the most expensive part of a SWIR spectrometer, yet we already have the FLIR TAU purchased and sitting idle. We plan on making the FLIR TAU a “removable” component of the SWIR spectrometer, so it can be swapped between the spectrometer and the eventual SWIR payload.
Build II: A Visible Transmission-Grating Hyperspectral Imager
In order to practice going through one cycle of optical-optomechanical design and tolerancing the system, we have been working on Build 2 (this project is currently in progress). We are making this using COTS optics and custom-machined mechanical housing, focusing on ease-of-use and disregarding volume constraints.
Nevertheless, one would be able to pick up this imager, carry it to Kings College Circle, and “scan the grass and measure its spectrum”. We will use Build 2 to try out various optical testing procedures that we’d use to eventually qualify the FINCH EYE, such as a spatial / spectral resolution test, SNR test, MTF test, stray light, etc.
We can enlist the help of Data Processing to create a GUI software that would seamlessly collect instrument data and process the hyperspectral data cube. Meanwhile, the Science team can find interesting samples for us to study with this instrument.
Build II: Flight Model
Satellite Earth-Observation instruments are frequently qualified on an airborne flight campaign. The instrument is mounted to a small commercial aircraft flown at low altitudes over the ground. The instrument can thus collect “realistic” data, which captures the effect of atmospheric attenuation, solar glare, motion blur, and atmospheric / cloud scattering. While the instrument would be extensively characterized in the lab regardless, an airborne flight campaign is an important milestone and one of several crowning achievements we can have as a team.
We will “fly” Build 2 on an aircraft, together with an “auxiliary” RGB camera that will take regular aerial images of the ground. Data Processing can work on some data fusion between the two, producing a geo-referenced image of the Earth together with the underlying hyperspectral datacube. The same software would be required again for the FINCH EYE payload once it is in outer space.
While Data Processing can acquire and collate data from both cameras using their manufacturer-provided API, the Firmware team can work on a board that collects data from the FLIR TAU which could be included as yet another auxiliary imager in the flight campaign. Using commercially-available SWIR objective lenses, we can construct a transmission-grating-based hyperspectral imager exactly as Build 2.
The Opto-mech team can develop a secure case for the instrument to attach to the aircraft. The Science team can plan a flight path over areas with scientific interest (e.g., farmland, river deltas, etc.)
Literature Review of Hyperspectral Imaging by Nanosatellites
After finishing with their crop residue work, the Science team could perform a literature review of hyperspectral imaging across UV-VIS-NIR-SWIR ranges. This would allow us to examine other potential use cases for the FINCH EYE payload and better contextualize our contributions within the broader efforts in academia and industry. This can be submitted in a conference / journal paper, especially with the emphasis on the advantages offered by nanosatellites, specifically low cost and fast revisit times in constellations.
Meanwhile, there are commercially-available hyperspectral camera systems that we can “evaluate”, by engaging the vendor as a prospective customer. After demo-ing a VNIR hyperspectral camera, we have the opportunity to trial their SWIR and VSWIR cameras. This would allow Science to enlarge their dataset and improve its quality. For example, Science can prepare a sample with soil and crop residue with varying degrees of soil moisture content and/or illuminated at different angles, and study these effects on the resulting spectral measurement. By adjusting the exposure time on the camera, one can control the image SNR, and hence we can perform a sensitivity analysis study to determine the minimum required SNR for accurate spectral unmixing.