CGNX Cognex Corporation

Research status: Primary-source verified through FY2025 Form 10-K plus company-issued Q2 2026 results. Valuation basis: approximately $60.4 in late August 2026; refresh before any capital decision.

Cognex is a high-quality machine-vision company positioned at the intersection of industrial automation, computer vision and physical AI. Its cameras, vision systems, barcode readers and software let machines inspect, identify, guide and measure objects in manufacturing/logistics environments. Q2 2026 results showed a strong cyclical/operating recovery: revenue was approximately $291M, +17% YoY, adjusted EBITDA margin about 32.2%, adjusted EPS $0.45, operating cash flow $69M and free cash flow $68M. At 5 July 2026 Cognex reported $755M of cash and investments and no debt. Full-year 2026 guidance calls for $1.13–1.15B revenue, 29–31% adjusted EBITDA margin and $1.64–1.68 adjusted EPS.

The attraction is a capital-light, debt-free physical-AI platform with strong gross margins and embedded engineering know-how. The challenge is that machine vision remains cyclical, exposed to factory automation and large project timing, while the current valuation already anticipates a meaningful earnings recovery.

PowerFund 24-month scenarios: Bear 20% = $47.5; Base 55% = $86.4; Bull 25% = $136.8; weighted $91.2, ~22.9% annualized. 60-month: Bear $52.8; Base $126; Bull $238; weighted $139.4, ~18.2% annualized. These are PowerFund scenarios, not company guidance or analyst targets.

Investment case

Cognex is one of the cleanest public-market ways to own machine perception for physical automation. As factories and warehouses become more autonomous, robots and production lines need reliable vision to identify parts, inspect defects, read codes, locate objects and make real-time decisions. Cognex combines optics, sensors, embedded compute, algorithms, software tools and a large application-engineering ecosystem.

The moat comes from decades of machine-vision data/know-how, ease of deployment, installed-base familiarity, channel/application expertise and the high cost of false positives/negatives in industrial environments. AI can expand the addressable market by making difficult vision tasks easier to configure, but it can also lower barriers if general-purpose vision models commoditize portions of the stack.

The thesis is investable if (1) 2026 recovery broadens beyond one or two end markets, (2) AI-enabled products such as OneVision improve win rates and ease of use without collapsing pricing, (3) adjusted EBITDA margins remain structurally near/above the high-20s as growth returns, and (4) the balance sheet is used rationally rather than masking weak organic growth with expensive M&A.

Verified operating baseline

• Q2 2026 revenue: approximately $291M, +17% YoY.
• Q2 adjusted EBITDA margin: approximately 32.2%.
• Q2 adjusted EPS: $0.45.
• Q2 operating cash flow: $69M, +60% YoY.
• Q2 FCF: $68M, +70% YoY; 93% of net income.
• Cash and investments: $755M; no debt at 5 July 2026.
• Q3 2026 guidance: revenue $300–320M, adjusted EBITDA margin 32–35%, adjusted EPS $0.50–0.54.
• FY2026 guidance: revenue $1.13–1.15B, adjusted EBITDA margin 29–31%, adjusted EPS $1.64–1.68.

Valuation scenarios

24 months

Weighted value: ~$91.2.

60 months

Weighted value: ~$139.4.

• Continued Q3/Q4 revenue acceleration and full-year margin expansion.
• AI-enabled vision products increasing ease of deployment and expanding addressable use cases.
• Recovery in consumer electronics, automotive, logistics and general industrial capex.
• Strong FCF conversion and disciplined use of the net-cash balance sheet.
• Higher recurring software/service content around the installed base.
• New large customer programs in logistics/warehouse automation and electronics.
• Evidence that AI lowers implementation cost faster than it lowers pricing.

• Industrial cyclicality: machine-vision spending can fall sharply when factory capex pauses.
• Customer/project timing: large electronics/logistics projects can distort growth.
• AI commoditization: general-purpose vision models and cheaper cameras/compute could reduce differentiation in easier applications.
• Competition: Keyence, SICK, Omron, Basler and numerous specialized vision vendors.
• Valuation: the recovery is increasingly reflected in the share price.
• China/geographic exposure: industrial demand and local competition can be volatile.
• Product execution: failure to translate AI research into easier-to-use commercial products could limit share gains.
• M&A: acquisitions can dilute the historically strong organic/capital-light model if poorly integrated.

WARNING / FREEZE ADDITIONS

• Organic revenue growth falls below ~5% for two quarters while industrial peers improve.
• Adjusted EBITDA margin falls below ~25% despite revenue growth.
• FCF conversion falls below ~80% of net income on a trailing-year basis without timing explanation.
• Large customer/project concentration rises materially.
• AI-enabled product launches fail to generate measurable adoption or pricing power.

REDUCE

• FY revenue/EPS expectations fall >10% from broad demand weakness or share loss.
• Gross/EBITDA margins contract for two consecutive quarters while Keyence/peer results remain strong.
• Net cash is deployed into large low-return acquisitions.
• China/local competition produces persistent pricing erosion.

EXIT / THESIS INVALIDATED

• Machine-vision differentiation becomes commoditized and Cognex cannot defend premium margins.
• Cognex structurally loses share in core high-end vision applications across multiple verticals.
• AI lowers industry barriers materially faster than Cognex can monetize its installed base/software/application expertise.
• The company abandons capital discipline and becomes an acquisition-led low-return industrial platform.

Cognex's toughest benchmark is Keyence, which combines excellent sensors/vision products with extraordinary direct-sales economics. SICK, Omron and Basler are meaningful competitors in vision/sensing and industrial automation. Cognex's advantage is specialization in machine vision, broad application know-how and an installed base across many production environments.

Physical AI expands the opportunity set but changes the competitive landscape. General foundation vision models can make classification easier, while industrial customers still require deterministic latency, ruggedness, calibration, traceability and extremely low error rates. Cognex wins if it packages modern AI into industrial-grade tools that non-experts can deploy reliably.

Portfolio interaction: CGNX would give PowerFund genuine Robotics & AI exposure with far less financial leverage than SYM and less healthcare-specific concentration than ISRG. It is more cyclical than ISRG but potentially offers stronger rerating upside if the factory-automation recovery continues.

1. Obtain/verify the Q2 2026 10-Q when indexed and reconcile company-issued non-GAAP metrics to GAAP results.
2. Break revenue growth by end market and geography; identify how much is cyclical recovery versus share gain.
3. Track OneVision/AI-enabled product adoption, average selling prices and gross-margin impact.
4. Compare Cognex growth/margins with Keyence, SICK and Basler on equivalent periods.
5. Analyze China exposure and local competitors, especially in lower-cost vision applications.
6. Build net-cash-adjusted valuation and FCF scenarios; avoid valuing the company only on recovering EPS.
7. Refresh 24/60-month scenarios after Q3 results; require a clear margin of safety versus ISRG/SYM alternatives.

Next review: Q3 2026 results or major AI-product/customer announcement.