We work on architectures, algorithms, protocols and tools that allow computer systems to self-adapt to their environment and self-organise to improve their performance, quality or ability to scale and survive. Self-generating Space Sensing Systems are lab and flight projects to which we are applying these concepts, particularly in the context of resource constrained ChipSat, Thin-Film and CubeSat scale spacecraft and subsystems, for earth observation and planetary science.
Adaptive communications and resource management at scale
An exciting new class of flight hardware has emerged allowing very low cost science, educational and commercial activities with large redundant arrays of inexpensive and disposable (RAID) thin-film, printed circuit board and CubeSat spacecraft.
Relying on efficient energy harvesting, these systems are very resource constrained, and require self optimizing algorithms to optimally manage low power peer-to-peer communications and to maximize the amount of data captured and reliably returned by sensors hosted by these spacecraft swarms.
We investigate applying and extending techniques we developed for low power meshes of terrestrial wireless sensor networks, reducing the power required to perform peer-to-peer spacecraft transfers at a range of 3km or greater, space-to-earth transfers at 350km or greater, and generally extending range while remaining within existing power budgets,
increasing and prioritizing sensing bit rates while staying within existing power budgets, and achieving reliable data collection from a swarm of individually unreliable spacecraft.
1g weather stations for earth and beyond
With collaborators at the University of Oxford and in industry, we are developing a one gram thin-film weather station suitable for collecting precision temperature, pressure and humidity data on earth, Mars and other bodies. An orbiting CubeSat locates the weather stations and forwards this information along with their beaconed meteorological data to a data warehouse on earth.
Parametric origami inspired thin-film systems
We have combined parametric subsystem generators with techniques our collaborators in aeronautics have developed for very lightweight adaptive aeronautics structures to create steerable deployable multifunctional small satellite subsystems with substantially better performance than current solutions.
These subsystems can provide up to 32x increase in current capabilities, simultaneously implementing any combination of the following while maintaining compatibility with standard small satellite interfaces:
- a deployable thin film solar array for an 1U CubeSat (initial proof of concept), capable of generating up to 128 W that stows in the same volume as a traditional 4 W array
- 64 cm diameter reflectarray antenna for high speed or long distance S-band communications
- aerodynamic drag deorbit device to satisfy international end of life deorbit / space debris regulations
The system is also applicable for high area to mass ratio instruments (e.g. magnetometer arrays), attitude control (e.g. magnetorquers), propulsion (e.g. solar sails) and large free flying thin-film spacecraft.
Spacecraft-on-Demand CPX scheduler
How can we optimise a network of manufacturing and sensor devices to collect scientifically relevant data at scales spanning hundreds of millions of kilometers and tens of years?
Spacecraft-on-Demand are very widely dispersed unreliable wireless sensor nodes with instantaneous mortality that can optimise the network of devices including informing the manufacturing devices that there is a need for replacement or augmentation. Optimisation strategies address hardware manufacture (including location and dispatch), and also their software, communications strategy and spatiotemporal configuration, during the data collection process adapting to a changing list of prioritised targets.
Spacecraft-on-Demand CPX is a project focusing on the self-adaptive scheduling aspects of this problem.