Overview
The robotics thesis is shifting from isolated industrial automation toward a broader “physical AI” stack: perception, embedded inference, motion control, sensing, power, communications and safety systems. Humanoid robots attract the headlines, but the more resilient investment area is the supplier layer that can sell into factories, logistics, medical devices, defence, autonomous vehicles, inspection robots and industrial equipment.
The page now integrates the uploaded Robotics Supply Chain Microcap Research report with an updated online check of company releases and market data. Treat market caps and prices as approximate snapshots: useful for screening, not for order entry.
ScalingTheme maturity
HighHardware execution burden
StrongPicks & shovels fit
SelectiveMicrocap investability
Stock Table
Working list from the robotics microcap research. The core ranking favours profitable or near-profitable enabling suppliers over speculative robot OEMs.
| Rank | Company | Ticker | Role in robotics stack | Approx. cap class | Research view |
|---|
| 1 | Ceva | CEVA | Edge AI / NPU / DSP / connectivity IP for smart-edge devices | Borderline microcap | Highest-quality “embedded intelligence” pick; royalty model, high gross margin, Physical AI optionality. |
| 2 | Vishay Precision Group | VPG | Force, strain, load-cell and precision measurement systems | Borderline microcap | Best force-sensing play; key bottleneck for contact-aware robotics and semiconductor equipment. |
| 3 | discoverIE Group | DSCV.L | Custom magnetics, sensing, controls and connectivity for industrial OEMs | Borderline microcap | Profitable customised electronics compounder; less pure-play, but high-quality design-in exposure. |
| 4 | SECO | IOT.MI | Computer-on-modules, edge computing, HMI, Clea IoT software | Direct microcap | Most orphaned edge-compute microcap; good thematic fit but requires execution discipline. |
| 5 | NCAB Group | NCAB.ST | High-mix, low-volume PCB sourcing for industrial, defence, medtech and AI infrastructure | Borderline microcap | PCB picks-and-shovels; strong Q1 2026 orders, but less robotics-specific. |
| 6 | Allient | ALNT | Precision motion, motors, controls and power solutions | Small-cap / borderline | Clean motion-control exposure; good company, but robotics is one of several end-markets. |
| 7 | Aeva Technologies | AEVA | 4D FMCW LiDAR and perception for autonomy, robotics and smart infrastructure | Borderline microcap | Interesting optionality, but still loss-making and dependent on production ramp. |
| 8 | Basler | BSL.DE | Industrial machine vision cameras | Borderline microcap | Profitable machine-vision exposure; cyclical recovery candidate. |
| 9 | Innoviz | INVZ | Automotive-grade LiDAR platforms and software | Direct microcap | High-risk LiDAR turnaround; Nasdaq compliance and cash-burn risk keep it outside the core list. |
| 10 | TT Electronics | TTG.L | Engineered electronics for medical, aerospace, defence and automation | Direct microcap | Turnaround/value situation, not a clean robotics leader yet. |
Value Chain Map
The robotics value chain is easiest to understand as a layered stack. The more upstream the company sits, the less dependent it is on one winning robot brand.
| Layer | What matters | Example suppliers | Investment note |
|---|
| Motion | Motors, drives, encoders, reducers, controllers, precision power | Allient, Harmonic Drive as reference | Core robotics bottleneck, but public microcap choice is limited. |
| Sensing | Force/torque, load cells, machine vision, LiDAR, tactile sensing | VPG, Basler, Aeva, Innoviz | Best mix of necessary function and supplier specialisation. |
| Embedded intelligence | Edge AI, DSP, NPU IP, COM modules, rugged boards, software stack | Ceva, SECO, Kontron as reference | High-quality angle because every physical AI system needs local compute. |
| Infrastructure | PCBs, magnetics, connectors, power systems, specialist electronics | discoverIE, NCAB, TT Electronics | Less pure-play, often more profitable and less dependent on robot hype. |
| Integration | System integrators, automation lines, deployment, maintenance, safety | Mostly larger industrials | Harder to find small listed pure plays; cyclicality matters. |
Technology Deep Dive
Robotics requires a closed loop: perceive the world, decide what to do, move precisely, verify contact, and keep operating safely. That loop explains why the best picks-and-shovels areas are sensing, motion and embedded compute.
| Technical bottleneck | Why it matters | Public-market angle |
|---|
| Force and tactile feedback | Robots that touch objects need to measure force, torque, strain and weight. Without this, manipulation stays brittle. | VPG is the clearest small listed force/strain sensing supplier. |
| On-device inference | Robots cannot send every perception and control decision to the cloud. Latency, reliability and privacy require edge compute. | Ceva and SECO give indirect exposure to edge AI hardware and software. |
| Motion control | Precision movement needs motors, drives, encoders and control systems working together. | Allient is a profitable US motion and controls supplier. |
| Perception | Robots need cameras, LiDAR, sensor fusion and robust detection in messy environments. | Basler, Aeva and Innoviz give exposure with very different risk profiles. |
| Electronics supply chain | Robotics adds electronic content: PCBs, magnetics, connectors, power distribution and rugged assemblies. | discoverIE and NCAB offer broader industrial picks-and-shovels exposure. |
Company Profiles
1. Ceva · CEVA
Edge AI / NPU IP · best quality score · borderline microcap
Ceva licenses silicon and software IP for smart-edge devices: NeuPro NPUs, SensPro DSPs, connectivity IP and 5G modem IP. Its robotics relevance is not that it builds robots, but that physical AI devices need efficient local inference and connectivity. Ceva said 2025 was a breakthrough year for AI licensing, with 10 NeuPro NPU agreements and AI contributing more than 20% of annual licensing revenue. It also reported record Q4 2025 revenue of $31.3m and FY2025 revenue of $109.6m.
- Why it is interesting: royalty/IP model, very high gross margin, NPU design wins, smart-edge exposure.
- Main risk: royalty ramp depends on licensee silicon shipping on time.
- Research rating: Core watchlist / highest-quality embedded intelligence pick.
2. Vishay Precision Group · VPG
Force, strain, load-cell and measurement systems · borderline microcap
VPG supplies the measurement layer: load cells, strain gages, force sensors and precision systems. This is important because practical robots need contact-aware sensing, especially in manipulation, semiconductor equipment, medical devices and industrial automation. VPG reported Q4 2025 revenue of $80.6m, up 10.9% year-on-year, and set 2026 objectives including mid-to-high single-digit revenue growth and higher bookings from initiatives.
- Why it is interesting: defensible measurement brands, physical AI/humanoid orders, semiconductor-test recovery angle.
- Main risk: industrial cyclicality and margin pressure.
- Research rating: Core watchlist / best sensing pick.
3. discoverIE Group · DSCV.L
Customised electronics, magnetics, controls, sensing and connectivity · UK-listed
discoverIE designs and manufactures customised electronic components for industrial OEMs. Its Q4 FY2026 trading update showed group orders up 16% at constant exchange rates and 15% organically, with strong demand in Magnetics & Controls and improving Sensing & Connectivity. It is not a pure robotics company, but it has the sort of design-in industrial electronics exposure that should benefit from rising electronic content in automation.
- Why it is interesting: profitable, recurring design-in relationships, exposure to industrial automation, medical, renewable energy and transport electrification.
- Main risk: less pure-play and dependent on industrial cycle.
- Research rating: High-quality infrastructure watchlist.
4. SECO · IOT.MI
Computer-on-modules, edge computing and Clea IoT platform · Italy-listed microcap
SECO supplies edge computing hardware, human-machine interfaces and IoT software. In February 2026 it reported unaudited FY2025 net sales of €200.7m at constant FX, up 9.4% year-on-year, with gross margin at 53.4%, exceeding guidance. The strategic point is simple: robotics, industrial automation and edge AI all need compute near the machine.
- Why it is interesting: direct edge-compute exposure, genuine microcap, hardware plus software stack.
- Main risk: execution, European small-cap liquidity and need to prove scalable profitability.
- Research rating: Orphan microcap watchlist.
5. NCAB Group · NCAB.ST
Advanced PCB sourcing · Sweden-listed
NCAB is an asset-light PCB supplier focused on reliable sourcing and delivery for demanding industrial customers. Q1 2026 net sales increased 12% to SEK 1,073.7m, order intake increased 27%, book-to-bill reached 1.20, and EBITA margin was 11.9%. It also cited strength in aerospace, defence, medtech and support systems for AI data centres.
- Why it is interesting: broad electronics bottleneck, strong order trends, defence/medtech/AI infrastructure exposure.
- Main risk: not robotics-specific and exposed to PCB capacity/pricing cycles.
- Research rating: Infrastructure watchlist.
6. Allient · ALNT
Motion, controls and power solutions · US-listed
Allient is one of the cleaner public-market ways to get profitable motion-control exposure. In Q1 2026, revenue rose to $138.9m, orders rose to $158.1m, book-to-bill was 1.14, and backlog reached $251.0m. Its presentation also highlighted industrial automation, vehicle demand, medical surgical robotics and aerospace/defence as relevant end-markets.
- Why it is interesting: real motion-control revenue, profitable, improved bookings.
- Main risk: robotics is one vertical among many, and the stock can trade like a cyclical industrial.
- Research rating: Solid motion-control watchlist.
7. Aeva Technologies · AEVA
4D FMCW LiDAR · high-risk autonomy optionality
Aeva is not a safe core holding, but it is one of the more interesting LiDAR names because its FMCW approach measures velocity and position together. Aeva reported FY2025 revenue of $18.1m, up from $9.1m, and guided 2026 revenue to $30m–$36m, implying roughly 70%–100% growth. It also cited a production contract with a top 10 passenger OEM, Nvidia DRIVE Hyperion reference sensor selection, Daimler Truck progress and a defence win with Forterra.
- Why it is interesting: strong commercial momentum and cross-market sensing applications.
- Main risk: operating losses, execution risk, production ramp risk.
- Research rating: Speculative optionality only.
Future Scenarios
Bull case: robots become a normal capital-equipment category. Edge AI, force sensing, machine vision and motion suppliers see expanding orders across multiple verticals.
Base case: industrial automation continues to recover, humanoids create demand signals but not yet mass revenue, and the best supplier names rerate selectively.
Bear case: robot deployments remain slower than the narrative, industrial capex weakens, LiDAR cash burn continues, and expensive supplier valuations compress.
Convergence
Robotics sits at the centre of several Futurology Hub themes:
- AI + Robotics: foundation models become physical-world agents.
- Robotics + Biotech: surgical robotics, lab automation, diagnostics and automated drug discovery workflows.
- Robotics + Defence: autonomous ground, air and underwater systems.
- Robotics + Mobility: autonomous vehicles, drones and delivery platforms.
- Robotics + Advanced Manufacturing: factory automation, inspection and flexible production.
Summary
The strongest way to invest in robotics from a microcap/picks-and-shovels perspective is to avoid the obvious robot hype and instead own the enabling layers: force sensing, embedded intelligence, motion control, machine vision and industrial electronics. The current priority watchlist is Ceva, VPG, discoverIE, SECO, NCAB and Allient. Aeva remains an interesting but speculative autonomy/perception option. Innoviz, TT Electronics and other turnarounds belong on a separate high-risk list until balance-sheet, listing, margin or profitability issues improve.
Current working conclusion: Robotics is investable, but selectively. The best opportunities are not necessarily the companies with the word “robot” in the pitch. They are the suppliers whose components become unavoidable if robots, automation systems and physical AI devices scale.