Reservoir Computing Research Sandpit: Apply for UK Defence Innovation Funding

If you’re a researcher working at the intersection of computational intelligence, physical computing, or defence applications, here’s an unusual opportunity that could change how you think about collaboration. UK Research and Innovation (UKRI) and the Defence Science and Technology Laboratory (Dstl) are running a four-day intensive “sandpit” event where researchers from different disciplines will come together to dream up innovative projects using reservoir computing for national security and defence.

This isn’t your typical grant application. You won’t submit a polished proposal and wait months for review. Instead, you’ll spend four days working face-to-face with people you’ve never met before, brainstorming solutions to real defence challenges. The best ideas developed during the sandpit could receive UKRI/Dstl funding to bring them to life. Think of it as speed dating for research collaboration, except instead of finding a partner, you’re building entire research teams and project concepts in real time.

Reservoir computing is a fascinating approach to machine learning that uses the inherent dynamics of physical or computational systems to process information. Unlike traditional neural networks that require extensive training of all components, reservoir computing uses a fixed, randomly connected “reservoir” and only trains the output layer. This makes it incredibly efficient and well-suited for edge computing, real-time processing, and hardware implementation—exactly the kind of capabilities defence applications need.

The deadline to apply is December 16, 2025 at 4:00 PM GMT, but this requires just an expression of interest at this stage, not a full proposal. If selected, you’ll attend the sandpit event where the real work begins.

Key Details at a Glance

What This Opportunity Offers

A Unique Collaboration Format: Sandpits are fundamentally different from traditional grant applications. You’re not applying with a ready-made proposal. Instead, you’re applying for the chance to participate in an intensive, facilitated workshop where you’ll meet researchers from completely different fields and work together to develop novel project ideas from scratch.

Access to Defence Expertise: Throughout the sandpit, you’ll have access to subject matter experts from Dstl who can explain real-world defence challenges and needs. This insider knowledge is invaluable because it helps you understand what problems actually matter to the defence community, not just what you think might be interesting theoretically.

Funding Potential: Projects that emerge from the sandpit and are selected for funding typically receive substantial support—often hundreds of thousands of pounds, depending on scope and duration. While there’s no guarantee, the sandpit format is designed to produce fundable ideas. The whole point is to create projects good enough to support.

Professional Network: Even if your specific project doesn’t receive funding, you’ll build relationships with researchers whose work complements yours. These connections often lead to future collaborations, publications, and other opportunities. Several past sandpit participants report that the networking alone was worth the time investment.

Skill Development: The sandpit process teaches you how to communicate across disciplines, think creatively under pressure, and develop research ideas collaboratively. These are increasingly important skills as research becomes more interdisciplinary and team-based.

Who Should Apply

This sandpit is looking for researchers who bring diverse perspectives to reservoir computing and its defence applications. You don’t need to be a reservoir computing expert already—in fact, diversity of expertise is the whole point.

Ideal Candidates Include:

• Computer scientists and AI researchers working on machine learning, neuromorphic computing, or unconventional computing paradigms
• Physicists and engineers developing physical reservoir systems using photonics, spintronics, mechanical systems, or other physical substrates
• Applied mathematicians studying dynamical systems, chaos theory, or nonlinear dynamics
• Neuroscientists exploring biological computation and brain-inspired algorithms
• Defence researchers with understanding of operational requirements, threat landscapes, or specific application domains
• Materials scientists developing novel materials with computational properties
• Electronic engineers working on hardware implementation of computing systems

What matters most is not your specific discipline but your ability to think creatively, work collaboratively, and contribute unique knowledge or perspectives to the team.

You’re a Strong Fit If You:

• Can articulate how your expertise might contribute to reservoir computing applications
• Are genuinely interested in defence and security applications (this isn’t just theoretical research)
• Feel comfortable working in teams with people from very different backgrounds
• Can commit to attending all sessions—both online preparatory meetings and the full in-person sandpit days
• Work at a UK research institution eligible for EPSRC funding (check UKRI’s institutional eligibility list)
• Can think on your feet and adapt ideas quickly based on group discussion

Not Sure If You’re Qualified? Apply anyway. The expression of interest stage is low-commitment, and the organizers are specifically looking for diverse expertise. If they think you’d contribute, they’ll invite you. If not, you’ve only spent an hour on the expression of interest—no harm done.

Understanding the Sandpit Format

If you’ve never participated in a research sandpit before, the format might seem unusual. Here’s how it typically works:

Pre-Sandpit Online Sessions: Before the main event, participants usually attend one or two online briefing sessions. These introduce you to the other participants, explain the defence context and challenges, and set up the framework for the in-person work. You’ll also likely receive background materials on reservoir computing applications and defence needs.

Day 1 - Introduction and Team Formation: The first day focuses on getting to know each other and understanding the problem space. Subject matter experts from Dstl present specific defence challenges where reservoir computing might help. Participants give short presentations on their expertise. By the end of day one, you’ll have a sense of who’s in the room and what problems might be tackled.

Day 2 - Idea Development: This is where things get intense. You’ll form small groups (usually 4-6 people from different disciplines) and start developing project concepts. Facilitators guide the process, but the ideas come from you. You’ll brainstorm, debate, sketch out approaches, and start to see which ideas have legs. Some groups merge; others split up and reform. It’s organized chaos, but productive chaos.

Day 3 - Project Refinement: Groups solidify around the most promising ideas and start fleshing them out. What would this project actually do? What expertise is needed? What would success look like? How would it advance both reservoir computing science and defence capabilities? You’ll start thinking about budgets, timelines, and deliverables.

Day 4 - Presentation and Selection: Each group prepares a short presentation of their project concept. You’ll present to the other participants and a panel that includes UKRI representatives and Dstl experts. They’ll ask tough questions about feasibility, novelty, and impact. Based on these presentations and subsequent discussion, certain projects may be invited to develop full proposals for funding.

Critical Point: You must attend all online sessions and all in-person days. This isn’t optional—the whole process depends on everyone being present throughout. If you can’t commit to the full schedule, don’t apply. The organizers need to know they can rely on participants to show up.

Insider Tips for a Successful Application and Experience

For the Expression of Interest: Even though it’s just an expression of interest, take it seriously. Explain clearly and concisely what expertise you bring and why you’re interested in defence applications of reservoir computing. Don’t write a research proposal—that’s not what they want. Instead, demonstrate that you understand what a sandpit is, that you have relevant expertise to contribute, and that you can work collaboratively.

Research the Topic Beforehand: Before the sandpit, read up on reservoir computing if it’s not your primary field. Understand the basic concepts—what reservoir computing is, how it differs from traditional neural networks, what its strengths and weaknesses are, and where it’s being applied currently. Similarly, think about defence and security challenges where computational efficiency, edge processing, or unconventional computing might matter.

Come Prepared to Listen: The most successful sandpit participants aren’t those who dominate conversation with their own ideas. They’re the ones who listen carefully to what others say, ask good questions, and help synthesize different perspectives into coherent project concepts. Your job is to contribute your expertise while being genuinely curious about what others bring to the table.

Think About Implementation: Defence funders care deeply about feasibility and transition to practice. Pure theoretical work is less interesting than approaches with clear paths to implementation and testing. As you develop ideas during the sandpit, keep asking: “How would we actually build this? How would we know if it works? How might it be deployed in operational contexts?”

Be Flexible About Roles: You might go into the sandpit thinking you’ll be the project lead on whatever emerges. But maybe you’ll end up as a key contributor on someone else’s project, or you’ll co-lead with people from different disciplines. Stay flexible. The best outcomes happen when people focus on creating great projects, not on securing specific roles for themselves.

Build Relationships: Make an effort to connect with other participants during breaks, meals, and social time. The formal sessions are important, but informal conversations often spark the best ideas. Plus, these are your future collaborators—you want them to enjoy working with you.

Prepare for Intensity: Four days of intensive collaboration is exhausting, especially for introverts. The days are long, the pace is fast, and you’re constantly thinking and interacting. Take care of yourself—get sleep, eat properly, and take breaks when you need them. You’ll do better work if you’re not running on fumes by day three.

Application Timeline and Process

December 16, 2025 (16:00 GMT): Expression of interest deadline. You’ll submit a relatively short form explaining your background, expertise, and interest in participating.

Late December 2025/Early January 2026 (estimated): Invitations sent to selected participants. The organizers aim for a diverse group that brings complementary expertise. They’re looking for about 20-30 participants typically.

January/February 2026 (estimated): Online preparatory sessions. These might be scheduled once the participant group is finalized.

Dates TBD: Four-day in-person sandpit event. The exact dates aren’t specified yet but will be communicated to selected participants. You’ll need to keep this time completely clear—no double-booking with conferences or teaching commitments.

Post-Sandpit: If your project is selected for potential funding, you’ll work with your team to develop a full proposal, usually within a few weeks. This proposal will go through a standard peer review process.

Application Materials Needed: The expression of interest typically requires:

• Your name, position, and institution
• Brief CV or biography (often limited to one page)
• Statement explaining your relevant expertise (usually 500-1000 words maximum)
• Description of why you’re interested in this sandpit (another 500 words or so)
• Confirmation that you can attend all required sessions
• Your line manager’s approval or confirmation that you can participate

What Makes a Strong Participant

Based on past UKRI sandpits, successful participants typically demonstrate:

Intellectual Flexibility (40% of selection criteria): Can you work outside your comfort zone? Can you contribute to projects that might not be exactly what you’d propose on your own? The best sandpit participants are comfortable with ambiguity and can pivot between different approaches as the group discussion evolves.

Relevant Expertise (30%): You need to bring something substantive to the table. This could be deep knowledge of reservoir computing, understanding of specific defence domains, expertise in hardware implementation, experience with machine learning applications—whatever it is, it should be clear how your knowledge contributes to the sandpit’s goals.

Collaboration Skills (20%): Can you work effectively in teams? Can you communicate your ideas clearly to non-specialists? Do you listen well and build on others’ contributions? The organizers will look for evidence of past collaborative work and your explicit statement about why collaborative research appeals to you.

Commitment to Defence Applications (10%): This isn’t pure academic research for its own sake. Projects need to address real defence and security needs. You should genuinely care about these applications and understand why they matter. If defence work makes you uncomfortable, this probably isn’t the right opportunity.

Reservoir Computing for Defence: What Problems Might You Tackle?

To give you a sense of what the sandpit might explore, here are some potential application areas where reservoir computing could address defence needs:

Edge Intelligence for Autonomous Systems: Drones, unmanned vehicles, and other autonomous platforms need to make fast decisions with limited power and computational resources. Reservoir computing’s efficiency makes it attractive for edge deployment where traditional AI models are too power-hungry or slow.

Secure Communications: Physical reservoir systems might offer new approaches to encrypted communications or signal processing that are resistant to certain types of attacks.

Radar and Sensor Processing: Processing sensor data in real-time to detect threats, classify objects, or predict patterns is fundamental to defence. Reservoir computing might enable faster, more efficient signal processing.

Anomaly Detection: Identifying unusual patterns in network traffic, sensor data, or system behavior is critical for cybersecurity and threat detection. Reservoir computing excels at time-series prediction and anomaly detection tasks.

Adaptive Camouflage or Countermeasures: Systems that need to adapt quickly to changing environments might benefit from reservoir computing’s fast training and adaptation capabilities.

Common Concerns and FAQs

Do I need security clearance? No, not for the sandpit itself. However, if projects move forward into implementation, some team members might eventually need clearance depending on the specific application. The initial sandpit and research work don’t require clearance.

What if I’ve never worked on defence projects before? That’s fine—many participants won’t have defence backgrounds. Dstl will provide context and expertise on defence needs. You’re being invited for your technical expertise, not defence experience.

Can postdocs or PhD students apply? Yes, as long as your institution is EPSRC-eligible. However, most participants tend to be early-career researchers, postdocs, or established academics. PhD students are less common but not excluded if they have relevant expertise.

What if reservoir computing isn’t my main research area? Apply anyway if you have relevant adjacent expertise. The organisers want diverse perspectives—someone who works on neuromorphic hardware or dynamical systems might contribute just as much as a reservoir computing specialist.

Is attendance really mandatory for all sessions? Yes, absolutely. The collaborative process breaks down if people drop in and out. If you can’t commit to the full schedule, wait for a different opportunity.

What happens if my project isn’t selected for funding? You’ll still have made valuable connections and learned about defence applications of your research. Many participants form long-term collaborations that lead to other projects, even if the specific sandpit proposal doesn’t get funded.

How much funding might projects receive? This varies widely depending on project scope, duration, and requirements. UKRI/Dstl partnerships typically fund projects at anywhere from £200,000 to £1 million+ over 2-4 years. The organizers will provide guidance during the sandpit on realistic budget ranges for different types of projects.

Can I participate remotely instead of attending in-person? The online sessions will be remote, but the four main sandpit days are in-person. Remote participation defeats the purpose of the sandpit format, which relies on intensive face-to-face interaction. If you can’t travel to the venue, this isn’t the right opportunity.

How to Apply

Ready to participate? Here’s your action plan:

First, verify that your institution is eligible for EPSRC funding. Check the UKRI website’s list of eligible research organizations. Most UK universities are eligible, as are many research institutes, but it’s worth confirming before you spend time on the application.

Second, talk to your line manager or supervisor. Make sure they’ll support your participation if you’re selected. You’ll need their approval, and you’ll need 4-5 days completely clear of other obligations. Get buy-in early.

Third, draft your expression of interest. Focus on:

• What specific expertise you bring (be concrete, not generic)
• Why reservoir computing for defence interests you (genuine enthusiasm beats generic statements)
• Evidence that you work well collaboratively (cite past team projects, interdisciplinary work, etc.)
• Confirmation that you understand the sandpit format and can commit to full attendance

Fourth, submit your expression of interest through the UKRI opportunity portal before 4:00 PM GMT on December 16, 2025. Don’t wait until the deadline—systems can be slow, and you want time to fix any technical issues.

Finally, if you’re not selected, don’t take it personally. Selection is as much about creating a balanced, complementary team as about individual qualifications. Keep an eye out for future sandpits on different topics.

Visit the official opportunity page for the application portal, detailed guidance, and any updates to dates or requirements: https://www.ukri.org/opportunity/sandpit-reservoir-computing-for-national-security-and-defence/

Questions about eligibility, the sandpit format, or application requirements? Contact EPSRC through the contact information provided on the opportunity page. They’re generally responsive to queries and happy to clarify whether you’d be a good fit.

This is an unusual opportunity to work on challenging problems with smart people from different fields, all while potentially securing funding for innovative research. If the format appeals to you and you have relevant expertise to contribute, it’s well worth the relatively small effort required for the expression of interest.