Space investment hits $7.95B. One million actuators. $1B for industrial robots.

Q1 2026 space investment hit $7.95B — nearly double the prior quarter — with the 12-month trailing total reaching a record $18.8B and average deal size climbing from $35.1M to $68M.

The largest Q1 deal was Saronic's $1.75B round, one of the biggest single space financings on record; SpaceX's confidential S-1 IPO filing (April 1, targeting $1.75–2T valuation) is amplifying sector-wide investor interest ahead of a June 2026 Nasdaq listing.

In Europe, VC investment in space rose 13% to €1.2B in 2025, but only 69% of rounds were European-led — and not a single growth-stage round was led by a European private investor, with public entities and US firms filling the gap.

SpaceX's IPO at $1.75–2T would be the largest in history, creating a sector-wide valuation repricing — every private space company will be benchmarked against a public SpaceX multiple, compressing or expanding valuations depending on strategic proximity.

The European financing gap is now a sovereignty question: one-third of European space company acquisitions since 2014 went to foreign buyers, and companies increasingly raise from US investors "out of necessity and not out of choice," per ESPI's lead analyst.

Investment is diversifying beyond traditional satellite communications into in-space infrastructure — stations, orbital data centers, and space-based manufacturing — reflecting a broadening of the sector's addressable market.

The SpaceX IPO will force a repricing of the entire enabling stack — launch services, satellite operators, ground station networks, and space domain awareness companies — creating entry points at both the infrastructure and application layers.

Europe's structural financing gap represents a direct opportunity for non-European LPs willing to lead growth rounds: European space companies with proven technology are actively seeking lead investors that European VCs currently cannot provide.

Risk caveat: SpaceX's valuation at $1.75–2T implies heroic revenue assumptions — Starlink, Colossus orbital data centers, and Starship all executing simultaneously. A miss on any segment compresses the sector-wide multiple.

Humanoid and Schaeffler signed a binding, phased deployment and supply agreement (May 13) to put 1,000–2,000 wheeled humanoid robots across Schaeffler's global manufacturing sites by 2032.

Initial deployment runs December 2026 – June 2027 across two German sites: Herzogenaurach (box handling in live production) and Schweinfurt (capability demonstration and full-scale integration testing).

Schaeffler simultaneously becomes Humanoid's preferred actuator supplier under a 5-year agreement covering 7-digit units (1M+) of joint actuators through 2031 — giving Schaeffler both robot customer and hardware supplier positions.

This is the largest binding humanoid deployment commitment disclosed to date — moving from "pilots" to contract-backed phased rollout with defined volumes, timelines, and integration requirements across a global industrial OEM.

The dual structure (deployment contract + actuator supply agreement) is the most sophisticated commercial arrangement in humanoid robotics yet: Schaeffler pays for deployed robots and sells the components that make them move — aligned incentives on both sides of the equation.

Humanoid was founded in 2024 by Artem Sokolov and is barely two years old. The speed from founding to a 2,000-unit binding contract with a Fortune 500-equivalent industrial supplier is without precedent in the humanoid sector.

The deal validates the Robot-as-a-Service model at industrial scale: Humanoid provides fleet management, maintenance, software updates, and 24/7 support — recurring revenue architecture rather than one-time hardware sales.

The actuator supply agreement creates a second investable angle: Schaeffler's planetary gear actuators (25–30 per humanoid, torque range 60–250 Nm) are now a confirmed input to a 1M+ unit order pipeline — the humanoid actuator supply chain is no longer theoretical.

Risk caveat: the initial phase is still beta-stage robots focusing on validation rather than full commercial operation. The path from December 2026 pilot to 2,000 deployed units by 2032 depends on performance milestones the agreement requires Humanoid to meet.

Mind Robotics raised $400M led by Kleiner Perkins, with new investors Meritech Capital, Redpoint Ventures, SV Angel, and VW Ventures, alongside returning backers a16z, Accel, Eclipse, and Bain Capital Ventures.

The round values Mind Robotics at $3.4B — up from $2B in March — making it one of the fastest valuation ramps in industrial robotics history: $0 to $3.4B in under six months.

Rivian is both a shareholder and operating partner, giving Mind Robotics access to a live, high-volume EV factory for training and deploying its models — a data flywheel most robotics startups cannot access.

Scaringe's thesis is that manufacturing robotics cannot be solved with software alone or hardware alone — it requires full-stack ownership: foundation model, hardware, and deployment infrastructure in the same company.

The Rivian factory access is genuinely differentiated: most robotics companies train in simulation or limited pilots. Mind Robotics trains on real manufacturing tasks at real production volume — the kind of environment that actually closes the sim-to-real gap.

The VW Ventures participation is strategically significant — VW already has a software joint venture with Rivian, meaning Mind Robotics now sits at the intersection of two of the world's largest EV manufacturers' supply chains.

Three rounds in six months with a16z, Kleiner, Accel, and VW all at the table signals that top-tier institutional investors see this as a generational bet — comparable in framing to early Waymo or early Scale AI.

Scaringe's framing of Mind Robotics as a US manufacturing competitiveness play — not just a robotics company — gives it a political and policy tailwind that most deep-tech startups lack.

Risk caveat: no commercial revenue has been disclosed and Rivian itself remains loss-making. The entire thesis rests on robots eventually performing reasoning-intensive dexterous tasks at production scale — a capability gap that current systems, including Mind's, have not yet fully closed.

The ICVP's third participant was successfully implanted at Rush University Medical Center (May 7) — with zero device failures across all three implantations, confirming the system's surgical reliability at multiple sites.

The device places 34 fully wireless stimulators with 544 electrodes directly into the visual cortex — operating independently of the retina and optic nerve, which are damaged or absent in the target patient population.

The system represents nearly 30 years of research by Prof. Philip Troyk at Illinois Tech, with the program now advancing to active 4-week post-implantation recovery and stimulation testing phases.

The ICVP targets patients with complete blindness from eye disease or trauma — a population that cannot benefit from retinal implants (the dominant existing approach) because their damage is proximal to or within the eye itself, making this a genuinely new patient category.

Three consecutive implantations without device failure is the key signal: it demonstrates the system is reproducibly safe across patients, a prerequisite for FDA's pathway from feasibility study to pivotal trial.

This is distinct from every BCI story we've covered recently — Motif (depression), CorTec (stroke rehabilitation), and INBRAIN (brain mapping) are all motor or psychiatric indications. ICVP is sensory restoration: giving someone the ability to perceive the world, not just interact with it.

The ICVP program is academic-led (Illinois Tech, 30 years of NIH/DARPA funding) rather than venture-backed — but its clinical data directly de-risks the investment case for private cortical visual prosthesis companies that are commercializing adjacent approaches (Second Sight, Pixium Vision successors).

Successful visual cortex stimulation also validates the broader cortical mapping market: the same intracortical stimulator array technology has applications in motor rehabilitation, memory enhancement, and seizure control.

Risk caveat: this is a feasibility study, not a pivotal trial — the three implantations are proof of surgical safety, not evidence of vision quality or commercial viability. Participants are in early recovery phases and stimulation outcomes have not yet been reported.

Quantum Motion raised $160M Series C co-led by DCVC and Kembara, with new investors British Business Bank and Firgun, plus returning backers Bosch Ventures, Porsche SE, and Oxford Science Enterprises — the largest VC round for a UK quantum company on record.

The company builds silicon spin-qubit processors using standard 300mm CMOS wafer technology via GlobalFoundries — enabling quantum systems that fit inside standard data center racks at claimed 100-fold cost reductions and 1,000-fold lower energy versus current approaches.

Quantum Motion deployed the first full-stack silicon quantum computer at the UK NQCC in 2025 and advanced to Stage B of DARPA's Quantum Benchmarking Initiative — moving from laboratory research into industrial evaluation.

The core thesis is that quantum computing will only scale if it uses the same manufacturing infrastructure as classical computing — silicon CMOS, not exotic cryogenic hardware. If the thesis holds, quantum computers could be deployed in existing data centers rather than requiring dedicated facilities.

Silicon-based quantum fits directly into the AI power story: as AI data centers strain grids, quantum computing optimized for energy efficiency and standard rack form factors arrives precisely when hyperscalers are most sensitive to power and space costs.

The British Business Bank's participation is a signal that UK government sees Quantum Motion as a strategic asset — fitting the UK's £2B quantum infrastructure commitment announced in March 2026.

Quantum Motion is privately held and the Series C is explicitly commercialization capital — not additional research funding. The company is building toward utility-scale systems using a manufacturing pathway that already exists(GlobalFoundries), which de-risks the production bottleneck that plagues most quantum competitors.

The DCVC lead is notable: DCVC is the firm behind deep-tech plays in aerospace, defense, and infrastructure — their quantum thesis is not consumer tech but industrial and national security computing.

Risk caveat: silicon spin qubits face their own coherence time and gate fidelity challenges that differ from superconducting or trapped-ion approaches. Quantum Motion's NQCC system is a proof of concept, not a commercially viable system yet — the claim of 100-fold cost reduction is directional, not demonstrated at scale.

Helion is building Tiny Merge, a fusion testbed one-eighth the size of its Polaris prototype, designed to iterate on plasma ring formation physics faster and at lower cost than full-scale machines — targeting operation by end of summer 2026.

The company has under three years to meet a contractual obligation to Microsoft (and partner Constellation) to deliver commercial fusion electricity, with financial penalties if the 2028 deadline is missed.

Tiny Merge's first tests will focus on plasma formation and merging — processes Helion has studied in previous prototypes but where "new results have prompted further questions," per director of electrical engineering Manav Singh.

Helion's decision to build smaller before scaling up is a significant strategic pivot — the entire industry logic has been to build progressively larger machines. Going smaller signals that fundamental plasma physics questions remain unresolved, which matters for any investor assessing the 2028 timeline credibility.

The Microsoft deadline is unique in fusion: no other fusion company faces a hard contractual date with financial penalties. This creates a verifiable binary event — by 2028, Helion either delivers net energy from fusion or it doesn't — making it the most trackable milestone in the entire fusion sector.

Microsoft's motivation for the 2028 deadline is AI data center power demand — the same infrastructure constraint driving investment in CFS, TerraPower, and OPG Darlington. Fusion's route to commercial power is increasingly tied to hyperscaler timelines rather than utility deployment.

Helion is privately held with $1B+ raised (investors include Sam Altman and OpenAI). A successful 2028 delivery would validate not just Helion but the entire private fusion sector's commercialization thesis — and a miss would reprice risk across the category.

The Tiny Merge approach — smaller, faster, cheaper iteration cycles — is itself an investable framework: companies that can compress the physics-to-engineering loop will reach commercial fusion earlier and at lower capital cost than those locked into sequential large-scale builds.

Risk caveat: Helion's approach (field-reversed configuration, magneto-inertial fusion) is less scientifically mature than tokamak or stellarator approaches. The new plasma questions Tiny Merge is designed to answer are foundational — if they're not resolved by late 2026, the 2028 deadline becomes very difficult to meet.

Physical AI startups raised a record $27.6B in 2025 — more than doubling year-on-year

Figure AI's $39B valuation and Skild AI's $14B valuation within months of each other signal institutional-scale conviction

Morgan Stanley projects over 1 billion humanoid robots in operation by 2050, generating a $5T market

Actuators and motion systems account for 40–50% of humanoid unit cost — the most accessible picks-and-shovels play in the category