Robotics
PrivateThe EKSO equity no longer exists as a medical-robotics bet — it became CHRN, an Applied Digital GPU-compute shell (~97% APLD-owned). The historical Ekso Bionics is a textbook sub-scale medtech that never cleared escape velocity ($18M peak revenue, perpetual losses, 1:15 reverse split, near-zero stock) and sold its listing to a datacenter roll-up. POST-MORTEM: dead as a standalone; any live "EKSO" exposure is now a leveraged AI-datacenter proxy, not exoskeletons.
Research
The verdict
The EKSO equity no longer exists as a medical-robotics bet — it became CHRN, an Applied Digital GPU-compute shell (~97% APLD-owned). The historical Ekso Bionics is a textbook sub-scale medtech that never cleared escape velocity ($18M peak revenue, perpetual losses, 1:15 reverse split, near-zero stock) and sold its listing to a datacenter roll-up. POST-MORTEM: dead as a standalone; any live "EKSO" exposure is now a leveraged AI-datacenter proxy, not exoskeletons.
Ekso Bionics Holdings was a Richmond/San Rafael, California robotic-exoskeleton company (founded 2005 as Berkeley Bionics; public via reverse merger 2014) that designed, manufactured and serviced wearable powered exoskeletons. It ran two segments:
How it made money: capital sales and leases of devices to rehabilitation hospitals (the "Enterprise Health" channel — ~75–80% of near-term revenue), plus service contracts, extended warranties, and clinical-training/CEU programs; and, increasingly, personal-device sales to individuals funded by insurance/Medicare reimbursement ("Personal Health"). In late 2025 it named National Seating & Mobility (NSM) the exclusive U.S. Indego Personal distributor in the complex-rehab-technology channel. Customers = rehab hospitals, VA facilities, universities (e.g. Vanderbilt, Burke/Weill Cornell), and individual SCI/stroke/MS patients. Contract structure was transactional and lumpy — big-ticket ($70k–$150k/unit class) capital equipment with no meaningful recurring-revenue base, which is the structural root of its quarter-to-quarter volatility.
Bottom line on the model: a real, FDA-cleared, clinically-validated product with a genuine unmet-need story — trapped in a business that was too small and too lumpy to ever cover its own cost structure. Peak annual revenue of $18.3M (FY2023) against tens of millions in cumulative losses is the whole tragedy in two numbers.
Named-stakeholder map (mostly ``, as no supply-chain filing was ingested):
Chokepoints / single-source dependencies: (1) Reimbursement is the true bottleneck — the personal-device market only exists to the extent CMS/insurers pay; a coverage-policy change is an existential input, not a cost line. (2) FDA clearance gates every indication and every product iteration. (3) Concentrated institutional buyers with long, budget-cycle-driven purchasing = revenue you cannot forecast. This lens is deliberately specific: the exoskeleton BOM is commoditized; the scarce inputs are regulatory clearances and payer coverage, and Ekso controlled neither.
What Ekso actually had:
Why the moat was shallow in practice:
Verdict: a narrow, non-durable moat. Real IP and evidence, zero cost advantage, no network effect, no lock-in strong enough to outrun the cash burn. Moats that don't compound into unit economics are just nice press releases — and this one ended in a reverse merger.
No segments.csv data was ingested (empty scaffold), so segment splits are `` and partial. Ekso reported EksoHealth vs EksoWorks in filings and, operationally, framed revenue as Enterprise Health vs Personal Health:
Trend & cause: total revenue decelerated then rolled over — FY2022 $12.9M → FY2023 $18.3M (+42%) → FY2024 $17.9M (−2%) → 2025 collapsed quarterly (see Lens 5). The FY2023 jump was partly the Indego (Parker Hannifin) acquisition broadening the line; FY2024's flat-to-down print showed the acquisition did not create durable growth. Geographic split was not cleanly sourced; U.S. was the primary market with international distribution secondary. Hard-requirement honesty: precise per-segment dollar splits are n/a (would require ingesting the 10-K segment footnote).
All figures `` from press releases (GlobeNewswire / StockTitan / Nasdaq / IR). The arc is a company shrinking into distress:
| Period | Revenue | YoY | Gross margin | Net loss | EPS |
|---|---|---|---|---|---|
| FY2022 | $12.9M | — | ~mid-40s% | large | — |
| FY2023 | $18.3M | +42% | ~49–50% | −$15.2M | — |
| Q4-2024 | $5.1M (record qtr) | +5% | ~53% | — | — |
| FY2024 | $17.9M | −2% | ~53% ($9.5M GP) | −$11.3M | −$0.56 |
| Q1-2025 | ~$3.4M (implied) | — | — | improving cash burn (−43%); +$3.8M warrant proceeds | — |
| Q2-2025 | $2.1M | −58% (vs $5.0M Q2-24) | — | −$2.7M | −$1.24 |
| Q3-2025 | $4.2M | +2% (vs $4.1M); +105% QoQ | 60.3% ($2.5M GP) | −$1.4M (−31% YoY) | — |
| Q1-2026 | $2.14M | −37% vs Q1-2025 | — | — | −$1.93 |
.
What the prints say:
No transcripts ingested; sentiment is `` from call recaps (Investing.com transcripts, Yahoo highlights). Tone arc across the last several calls:
Sentiment read: a classic distressed-microcap tonal drift — from "growth" → "mixed, disciplined" → "we have an AI story" → "we're selling the shell." Management leaned on narrative substitution (AI foundation model) as fundamentals deteriorated.
Peer set = pure-play medical/powered-exoskeleton names (the only honest comparables; the successor's true comps are datacenter/neocloud — see the overlay note). Multiples are `` or n/a; no multiple is fabricated.
| Company | Ticker | Mkt cap | TTM revenue | EV/Sales | EV/EBIT | P/E | Div yld | 5-yr avg ROE |
|---|---|---|---|---|---|---|---|---|
| Ekso Bionics (historical) | EKSO→CHRN | ~$25M pre-merger (est. distressed) | ~$14–18M | n/a | n/a — negative EBIT | n/a — negative EPS | 0% | negative |
| Lifeward (ReWalk) | LFWD | ~$10.2M (Nov 2025) | ~$24.5M (TTM Sep-2025) | n/a | n/a — negative | n/a — negative | 0% | negative |
| Cyberdyne | 7779.T | not sourced this run | not sourced | n/a | n/a | n/a | ~0% | negative (historically loss-making) |
| Ottobock | private | n/a — private | large (diversified O&P) | n/a — private | n/a | n/a | n/a | |
| DIH Medical | DHAI | not sourced this run | not sourced | n/a | n/a | n/a | 0% | negative |
Read: the entire listed pure-play exoskeleton cohort is sub-scale and unprofitable — Lifeward, the "largest," carries a ~$10M market cap on ~$24M revenue and negative earnings. This is a category, not just a company, problem: a decade in, no Western pure-play has reached self-funding scale. Ekso screened as one of the healthier-margin names but smallest-revenue — and still had to sell its listing. Multiples deliberately left n/a: with negative EBIT/EPS across the cohort, P/E and EV/EBIT are meaningless, and I did not source clean EV/Sales figures this run.
Mostly ``:
What the tape reveals about this name: for its final two years the market stopped reacting to exoskeleton fundamentals entirely and traded on (a) survival/dilution mechanics and (b) the AI-datacenter reverse-merger. The operating business became irrelevant to the share price well before it was formally sidelined — the equity was a corporate-action instrument, not a medtech.
``:
Assessment against the five tests:
Successor note: ChronoScale is run by Cenly Chen (CEO from 2026-05-06), an Applied Digital-side appointment — confirming the shell is now operated for the datacenter business, not the exoskeleton team.
No filings on the shelf, so income-statement/balance-sheet forensics are ``-level and directional, not line-item audited:
n/a — not verified without the 10-K footnotes).Regulatory findings (required sub-section):
regulatory/regulatory-findings.md records total_sec_findings: 0 and notes Ekso has cik: null in the index so no EDGAR enforcement search was run by the script. (Historically Ekso did file under CIK 0001549084; a future refresh could confirm directly, but no enforcement action is known.)"Ekso Bionics" (FTC OR DOJ OR FDA OR settlement OR fine OR penalty) enforcement surfaced no material enforcement action, consent decree, or penalty. FDA interactions are product clearances (510(k)), not adverse actions — i.e. regulatory approvals, the opposite of enforcement.n/a — 10-K not on shelf); the most-flagged "risk" in the FY2024 10-K per coverage was the merger/strategic-transaction risk itself, not litigation.A forward EPS projection for "Ekso Bionics" is not meaningful — the entity is gone. The legacy exoskeleton business is being divested from ChronoScale and, on a standalone basis, had no realistic path to positive EPS (sub-$20M revenue, structural losses). So this lens splits:
(a) The historical operating business — terminal, no forecast. Standalone Ekso EPS trajectory: negative, indefinitely. FY2024 −$0.56; 2025 quarterly EPS between −$1.24 and −$1.93 (post-split share counts). Base/bull/bear all converge on "does not reach profitability before running out of cash" — which is why it was sold. ``. No forecast.ts create — there is nothing scoreable here (and the mandate forbids it in watchlist mode).
(b) What a CHRN holder actually owns now — an Applied Digital datacenter proxy. Post-merger, "EKSO/CHRN" value is ~97% driven by Applied Digital's cloud/GPU-compute business, not exoskeletons. That business is a CoreWeave-anchored hyperscale-datacenter operator: at Ellendale, ND (Polaris Forge 1), APLD signed 250MW + 150MW = 400MW of CoreWeave leases with ~$11B total contracted lease revenue over ~15 years, financed partly by a $1.59B 7.0% senior secured notes (due 2031). So the CHRN thesis is a leveraged, single-tenant-concentrated (CoreWeave) AI-datacenter build-out — an entirely different risk profile (capex intensity, tenant concentration, financing/rate risk, execution timing on 2025–2027 ready-for-service dates) than a medtech.
Projection verdict: value the successor as a neocloud/datacenter proxy with heavy CoreWeave counterparty concentration and a levered balance sheet, not as an exoskeleton company. The legacy business is a rounding error / divestiture candidate. `` — a proper CHRN forecast belongs in a dedicated Applied Digital / ChronoScale deep-dive (recommended below), not here.
Bull case (for the historical exoskeleton business — now moot): real FDA-cleared products, first-mover clinical evidence, ~17% share, improving gross margin (60%+), a growing personal/Medicare-reimbursed channel, and an interesting AI foundation-model-for-human-motion optionality (NVIDIA Connect). In a world where the category inflects (the medical-exoskeleton TAM is quoted at ~$1.4B in 2025 → multi-billion by mid-2030s at ~17–19% CAGR ), a scaled leader could matter.
Bear case (which won): three permanent-impairment risks all fired — (1) structural sub-scale: revenue never crossed the threshold to cover a regulated-hardware cost base; (2) reimbursement/adoption gating: the personal market depends on payers, and adoption stayed too slow; (3) capital-markets dependence: chronic dilution + a going-concern balance sheet meant the equity was perpetually raising at lower prices, ending in a 1:15 reverse split and near-zero price. The market baked in terminal value ≈ 0 for the operating business.
Pre-mortem (it already happened): 18 months out, the thesis "Ekso scales its exoskeleton franchise" broke because the business was too small to self-fund and payer/adoption ramps were too slow; management, unable to grow into the cost base, substituted an AI narrative and then sold the listing to an AI-datacenter roll-up (Applied Digital), leaving legacy holders with ~3% of a completely different company.
Multiples too high? For the operating business, any positive equity value was too high relative to a zero-terminal-value standalone — which is why it was effectively given away. For the successor (CHRN), the risk inverts: you're now paying datacenter-buildout multiples on a levered, single-tenant-concentrated entity, and the exoskeleton legacy is noise.
Contrarian view (what the market refused to see, and then saw all at once): the market spent years pricing EKSO as a struggling-but-going medtech; the actual value event was a shell/listing transaction. The lesson: for terminal-stage microcaps, the equity can become a reverse-merger option on someone else's business — which is exactly what happened, and the ~$0.23 → ~$13.50 round-trip proves the operating fundamentals were never what mattered at the end.
Dismantling any residual "own EKSO/CHRN for the exoskeleton upside" bull case:
Directed at ChronoScale / Applied Digital leadership (Cenly Chen; APLD), since that is who now controls the listing — the legacy Ekso team is exiting:
The #1 knee/hip implant franchise priced for failure (~12x fwd EPS) — but it is the value trap until it proves organic growth can clear 3% without the Paragon/Monogram M&A crutch and stops losing the robotics war to Mako. Cheap is the thesis and the warning.
A cheap, well-run AIDC compounder mis-tagged "robotics" — it just SOLD its robots; the real bet is whether ~4% organic hardware growth + buybacks + a tariff-refund kicker re-rates a 13x stub the Street already targets at $330.
A near-breakeven Chinese smart-EV OEM whose margin (GM 18.9% FY25, ~20% Q1'26) and a high-margin VW software-licensing annuity are real — but FY26 volume has rolled over (-22.6% YTD), and the IRON/eVTOL/robotaxi "embodied-AI" optionality the bulls pay for is unproven cash-burn; long the software+margin inflection at a 52-week-low multiple, but only if the GX/new-model cycle re-accelerates deliveries by 2H26.