Design Challenges
On Crowdsourced Innovation
Crowdsourced design challenges are one of the more underappreciated tools in complex problem-solving. For organisations like NASA, they open a problem to a much wider pool of thinking than any internal team can offer - people who bring genuinely different mental models, reference points, and creative instincts. The solutions that emerge often have a freshness that comes precisely from not being constrained by domain familiarity. Sometimes it is the person who does not know what cannot be done who finds the way through.
For participants, the appeal is different but equally real. These challenges offer a chance to contribute to something meaningful, to learn about a topic you would never otherwise have cause to investigate, and to test your thinking against a real brief with real stakes. Some of my most substantial periods of self-directed learning have come from picking up a challenge in an unfamiliar area and working through what I needed to understand. Having to learn something from scratch in order to complete a task is one of the more effective ways to actually retain it.
Having competed in a number of these challenges, I now also design and run them - for the Roo-ver mission and the ALEPH project - with a perspective on what makes a challenge work. The framing matters enormously: a well-designed challenge gives participants enough context to engage meaningfully, a constraint set that is genuinely interesting to work within, and a problem that benefits from diverse approaches rather than one that simply requires prior expertise. Done well, it creates value on both sides.
Trusted Human-Machine Decision Making Framework
This challenge sought a framework for interaction between an automated system and front-line individuals that builds trust and confidence through communication, without compromising proficiency through burdensome exchanges. The context was high-consequence, real-time decision-making - situations involving safety of people, property, or services where humans must assess multiple information streams and act under pressure.
My solution comprised three elements: a UI framework for presenting recommendations in the context of information feeding into an autonomous system's trustworthiness assessment; parameters for an explainable model applied to existing AI systems; and a suggested meta-AI system for achieving that explainable model. On notifying me of the award, the challenge organisers noted that, of all submissions received - including from domain experts - mine was the only one that considered the human aspect of the problem.
NASA Lunar Delivery Challenge — Lightweight Inflatable Delivery System (LIDS)
NASA needed a practical, cost-effective system for unloading payloads from a range of commercial lunar landers at the Moon's south pole - varying mass, varying vehicle configurations, extreme environment. Current Earth-based logistics systems are too heavy to package for lunar deployment.
LIDS is an inflatable gantry with a novel rotatable crane, making the system compact, lightweight, and highly versatile. The system primarily consists of inflatable components to conserve weight and volume. It includes an autonomous attachment capability that allows it to connect to a wide range of lander hatch heights, and a rotatable component that acts as a forklift - combined with a built-in winch - enabling both vertical and horizontal payload movement.
NASA REALM Storyboard Challenge — RFID-Enabled Autonomous Logistics Management
REALM is a phased experimental project applying RFID technologies to automate logistics management for future space exploration - from inventory and item searches through to robotic interaction with cargo prior to human arrival. NASA needed a storyboard for a two-minute public-facing animation explaining the project and its significance.
My solution made the concept accessible by grounding it in familiar analogies from air travel, then connecting those to real applications aboard the ISS through illustrations that followed the REALM system across its three phases. The script and storyboard were later produced into an official NASA video.
NASA Foldable Radiation Shield Challenge
One of the major concerns for interplanetary travel is galactic cosmic radiation (GCR) - high-energy heavy elements that are difficult to shield against with traditional spacecraft architectures. NASA sought novel folding or origami-based concepts for packaging a large lithium hydride radiation shield within a single launch vehicle that could then be deployed around a Mars spacecraft in cislunar space.
My submission was an armadillo-inspired mechanism: overlapping shells that fold out to provide layered radiation protection, drawing directly from how armadillos use articulated keratinous plates to protect themselves. This was one of my earliest challenges - submitted before I had developed skills in 3D CAD, using the shapes tool in PowerPoint. Placing second was the starting point for a deeper exploration into 3D design, robotics, and 3D printing.
NASA ION Training Course — Interplanetary Overlay Network
ION is NASA's implementation of the Delay/Disruption Tolerant Networking (DTN) protocol - designed for communication in environments where signal delays and disruptions make standard internet protocols unworkable. The challenge was to design a modular five-day training course curriculum and develop an example module to demonstrate the format, depth, and approach.
This was the most demanding challenge I completed, because it required building understanding from nothing. Each layer of new concepts and acronyms revealed three more requiring investigation. That process put me in a position close to the learners I was designing for - and made it possible to construct something that builds knowledge progressively, using analogies, worked examples, and definitions that genuinely helped. I also proposed an alternative applied scenario: an IoN-based weather station placed in Jupiter's Great Red Spot, to help students appreciate the possibilities that follow from mastering the protocol. NASA purchased the infographic developed as part of the course materials.
NASA GrabCAD Regolith Sampling Challenge
NASA sought a lightweight, low-power sampling system capable of collecting and sorting lunar regolith for return to Earth - prioritising rocks in the 1-2 cm range while filling remaining volume with fines, all within tight mass, power, and thermal constraints. The system had to operate in hard vacuum across a temperature range of -140 to +80 degrees Celsius.
My design aimed at simplicity to reduce moving parts and maximise robustness. The sampler works by being moved along the ground, allowing only rocks under 2 cm to enter at the front. Collected material passes over a sifting grate separating 1-2 cm rocks from fines, with a weighing station determining when the minimum 1-2 cm rock quota has been met. The container rotates during depositing to create layered, spiralling distributions. Final mass: 0.354 kg. Power draw: 1.14 watts.
"The use of a mesh for filtering was interesting and well-received by the team. The openness of the design allows for great visibility during sample collection and filter operations. The detailed documentation, including a narrated video of testing, was highly appreciated by the team." NASA
NASA MarsXR Challenge — Extravehicular Activity Training Storyboard
This challenge sought storyboards for new assets and scenarios within NASA's Mars XR Operations Support System (XOSS) environment - training simulations for situations astronauts will encounter on Mars. The emphasis was on realistic, practically useful scenarios that prepare trainees for conditions they are likely to face.
My winning entry proposed a two-part scenario built around collaboration and unanticipated events. Part one involves unbogging a rover - a task requiring co-ordination between two astronauts through hand signals, shared problem-solving, and real-time adjustment. Part two begins immediately after: the fellow astronaut becomes unresponsive, and the agent must locate them, move them into recovery position, check vitals via a wrist monitor, and call for help. The scenario was designed to address a known issue in long-duration missions - the decline in collaborative and creative task performance over time - by training under compound, unexpected conditions.