> ## Content Index > Fetch the complete content index at: https://nexi.fund/llms.txt > Use this file to discover other available public pages before exploring further. # Defence Robotics' Capital Stack Is Splitting Into Buyers and Builders - URL: https://nexi.fund/defence-robotics-capital-stack-2026/ - Published: 2026-08-22T10:00:54.000Z - Updated: 2026-08-22T10:00:54.000Z - Description: Defence-robotics capital is concentrating into a two-tier stack: platforms buying scale through M&A at the top, governments seeding the base through grants, and standalone mid-stage firms caught in the squeeze. - Author: Nexi.fund Labs - Tags: Defence & Robotics, #mode-6, #hook-paradox, #track-A The capital that fled defence technology in 2022 is now writing its largest cheques of the decade. And it is splitting the market in two. ๐ŸŽฏ **Defence-robotics capital is concentrating into a two-tier stack.** At the top, platforms and primes are buying scale through acquisition. Ondas' $875.8M takeover of DZYNE and Kraken Robotics' C$615M grab of Covelya are the headline deals of the summer. At the base, governments are seeding the next layer with pre-institutional programmes. The UK's UKDI Novel Autonomy decisions landed on 20 August. The middle, where standalone startups once pitched themselves as the future, is being squeezed from both sides. #### Capital is pouring back in Defence-tech venture equity reached $49.1B in 2025, up from $27.2B a year earlier. ยท **Nexi Fund, 2026* #### The largest deal of the wave Ondas closed its $875.8M acquisition of autonomous-aircraft maker DZYNE in July. ยท **Breaking Defense, 2026* #### Scale in the water, too Kraken Robotics closed its C$615M takeover of subsea group Covelya in July. ยท **Nexi Fund, 2026* #### And a grant at the base The UKDI Novel Autonomy competition offers up to ยฃ600K per project to seed early-stage ideas. ยท **GOV.UK, 2026* ## The buyers are rewriting the cap table The top of the stack is where the money is loudest. Well-capitalised autonomy platforms are not building every capability in-house. They are buying the ones that already work. Ondas did it with DZYNE, folding long-endurance autonomous aircraft and counter-drone systems into a single operating platform. Kraken did it with Covelya, combining underwater sensors and navigation into one subsea business. Neither was a science project. Both targets were revenue-generating and mission-proven before the cheque was signed. The pattern is not limited to two deals. Agility Robotics went public through a $2.5B combination in June. Forterra secured a $92M production award alongside a prime to put autonomy on ground vehicles. A European prime folded a drone maker into its portfolio. Anduril, already a platform, kept absorbing specialist pieces โ€” radar, command-and-control, propulsion โ€” to close gaps in its own architecture. The logic is operational, not financial. A systems-of-systems builder wins by owning the interfaces between sensing, effect, and software. Acquiring a proven subsystem is faster than incubating one, and procurement offices trust fielded hardware over slideware. Global defence spending passed $2.63T in 2025\. The autonomy layer on top of that spend is where the buyers are placing their bets. ## The generalists are losing the room Not every robotics segment is riding the wave. The capital concentrating into defence is leaving the broad commercial market thinner. Logistics and warehousing robotics, for years the largest recipient of robotics venture funding, shed 28.5% of its deal value in 2025\. The e-commerce surge that funded it matured faster than expected, and the same funds now rotating into autonomy are rotating out of fulfilment centres. This is the paradox at the centre of the 2026 market. The robotics story did not get smaller. It got narrower. Capital that once spread across every warehouse and hospital corridor is now betting on systems that can operate where the stakes are physical and the budgets are sovereign. For a founder, the implication is uncomfortable. A compelling demo is no longer enough. The buyers at the top want evidence of deployment, and the governments at the base want a credible path to fielding. The middle โ€” the company that is interesting but not yet integral โ€” is the hardest place to raise. ## The incubators are planting the base While the top consolidates, the base is being seeded on purpose. Governments have decided that waiting for the market to produce the next generation is too slow. The UK's UKDI Novel Autonomy and Robotics programme, run under the Defence Science and Technology Laboratory's Autonomy Programme Incubator, closed Stage 1 submissions on 11 August. Decisions were released on 20 August. Each project can draw up to ยฃ600K, starting at technology readiness level (TRL) 1 or 2, with a mandate to advance at least one TRL level during the work. This is not a prize for finished products. It is pre-institutional capital for ideas that are years from the field. The explicit goal is to find cutting-edge, low-TRL science that might mature into military robotics across land, sea, and air. The same shape appears elsewhere. Energy and defence agencies run analogous grant and challenge programmes for autonomy. The mechanism is consistent: de-risk the earliest research so that, in three to five years, there is a deeper bench of companies for the buyers at the top to acquire. The base and the top are not separate markets. They are one pipeline, viewed at different altitudes. ## Two models, one pipeline | Parameter | Buyers (M&A) | Builders (incubation) | | ------------------- | ---------------------------------------------- | -------------------------------------- | | **Capital source** | โœ” Corporate balance sheets, VC, public markets | โœ— Government grants, agency programmes | | **Target maturity** | โœ” Revenue-generating, fielded systems | โœ— TRL 1โ€“2 research concepts | | **Goal** | โœ” Fill a platform architecture gap now | โœ— Seed the next bench of targets | | **Risk** | โœ— Integration, valuation, execution | โœ— Most ideas never graduate | Source: Nexi Fund and GOV.UK, 2026 ## Why the timing, not the thesis, changed The argument that autonomy would reshape defence is old. What is new is who believes it with a chequebook. For most of the last decade, large funds treated defence as an ethical exclusion. The thesis was already there โ€” cheaper autonomous systems versus billion-dollar platforms โ€” but the capital was not. That changed when operational buyers stopped treating autonomy as experimental and started treating it as inventory. Procurement offices have shortened their timelines. Fast-lane acquisition pathways, direct awards, and prime-aligned production contracts have turned "interesting" into "ordered." When the buyer at the top can commit revenue, the acquirer at the top can underwrite a takeover with confidence. The re-rating is not about one budget line. It is a structural view that autonomy is now core infrastructure for defence, not a future option. That view is what moved $49.1B into the sector in a single year. ## Integration is the real test The top of the stack is efficient until it is not. Every acquisition adds a culture, a codebase, and a supply chain that must merge without breaking the fielded product. Ondas is folding DZYNE in alongside other recent deals. Kraken is absorbing six Covelya subsidiaries across four continents. The cost synergies are real on paper โ€” Kraken targets C$10M within two years โ€” but integration is where acquisition theses go to die. The promised-versus-reality gap shows up here. A combined platform looks powerful in the press release. It looks different six quarters later, when two engineering organisations are still reconciling roadmaps and a customer is waiting on delivery. Investors are already nervous about valuation. One subsea acquirer saw its shares rally hard into the deal, then draw a downgrade citing exactly this risk. The market will pay for scale. It is less willing to wait forever for the proof. ## For founders, M&A is the new exit The listing window that opened for robotics in earlier cycles has narrowed. Agility's $2.5B public combination is the exception that proves the rule. Most autonomy founders now model their outcome as a sale to a platform, not a ticker. That changes how a company is built. A business optimised for acquisition looks different from one optimised for independence. It concentrates on a defensible subsystem, drives to a fielded proof point, and keeps its architecture clean enough for a buyer to absorb without a rebuild. The government programmes at the base are becoming the first customers that make those proofs credible. A founder who lands a UKDI grant is not just funded. They are benchmarked against a procurement standard that an acquirer at the top already trusts. The pipeline is closing the loop. Grants create the bench. Buyers acquire the survivors. The founder's task is to be the one both tiers want before the middle squeezes shut. ## What to watch ๐Ÿ“Š **Key signals to track** Follow-on M&A velocity โ€” whether Ondas and Kraken are the start of a broader rollup or the peak of pricing. UKDI Phase 2 awards โ€” the ยฃ600K Stage 1 grants are explicitly a feeder to larger follow-on funding. Conversion, not valuation โ€” whether the $49.1B in defence-tech VC reaches fielded systems or settles as paper. The warehousing decline โ€” whether commercial robotics contraction deepens or stabilises. ## The squeeze in the middle The two-tier stack is efficient. It funds the proven and the proto-scientific, and it lets primes scale without rebuilding. But it leaves a wound in the centre. A defence-robotics company that is too mature to win a grant and too small to be acquired sits in the worst position of the cycle. It must reach deployment on its own, against buyers who would rather buy than build and governments who only fund the earliest or the latest stage. That is the promised-versus-reality tension of 2026\. The capital is historic. The path from a good company to a bought company has rarely been narrower. [Novel Autonomy and Robotics (Phase 1)The UK Defence Innovation competition seeding early-stage autonomous-systems research, with Stage 1 decisions released 20 August 2026.GOV.UK (UK Defence Innovation / Dstl)](https://www.gov.uk/government/publications/novel-autonomy-and-robotics-phase-1?ref=nexi.fund) Editorial note: the primary source for the bottom-tier incubation signal. [Ondas makes $876M acquisition of DZYNE, latest in acquisition spreeThe headline top-tier deal folding long-endurance autonomous aircraft and counter-UAS into one platform.Breaking Defense](https://breakingdefense.com/2026/07/ondas-makes-876m-acquisition-of-dyzne-latest-in-acquisition-spree?ref=nexi.fund) Editorial note: the anchor transaction for the buyers tier. [The M&A Wave Reshaping Defence RoboticsMacro data on the $49.1B defence-tech VC year and the rollup across Ondas, Kraken, and peers.Nexi Fund](https://nexi.fund/defence-robotics-consolidation-2026/) Editorial note: the source for the capital-flow and segment-decline figures.