PM Portfolio Manager Reports

From Portfolio Manager reports

Humanoid OEM Platforms And Regional Ecosystems

This node covers finished humanoid robot makers and platform operators: the companies that assemble the body, sensors, batteries, compute, control software, teleoperation tools, fleet logs, service model, and regional supply chain into a usable robot. The money path is paid units, leases, Robots-as-a-Service contracts, service attach, software updates, and lower support cost that turns fleet data into margin and cash flow. The current public basket is watch-only: TSLA is the cleanest public humanoid platform route, SERV has the strongest current public fleet evidence, RR is the more direct robot-platform/RaaS proxy with stale filing risk, and PDYN and KSCP are lower-conviction autonomy or robot-fleet monitors.

Current Setup

Watch-only platform layer Finished platforms may own the largest economics, but public exposure is mostly proxy exposure.

Humanoid OEM platforms combine robot hardware, operating software, support tools, regional sourcing, customer task roadmaps, and fleet data. The proof is not a demo; it is paid fleet activity that shows utilization, support cost, margin, and cash conversion.

Direct public route TSLA is the only scaled public humanoid platform proxy in the ranked basket. Optimus, factories, autonomy data, AI infrastructure, and balance sheet matter, but robot revenue, unit cost, and task economics are not separately disclosed.
Fleet evidence SERV ranks ahead of RR because its fleet KPIs are fresher and more measurable. RR remains the more direct robot-platform/RaaS proxy, but its local lane still carries the delayed March 2026 10-Q caveat.
Regional gap China, Japan, Korea, and private U.S. leaders stay in the monitor list. Unitree, Figure, Agility, 1X, Apptronik, Boston Dynamics, UBTECH, Xiaomi, XPeng, Toyota, Honda, and Hyundai matter more to the map than to route-ready public cards.
  1. PrototypeRobot can perform bounded factory or service tasks
  2. Paid fleetCustomer pays for units, lease, RaaS, or service contract
  3. Fleet logsTask success rises and intervention rate falls
  4. Service modelSupport cost, uptime, repairs, and software attach improve
  5. EconomicsRevenue becomes margin, cash flow, and per-share value

Basket

The ranking is the final tournament merge of ten xhigh primary agents and one xhigh merge selector. It ranks source-backed platform exposure and operating evidence before pure thematic fit or chart timing. That moves SERV ahead of RR: SERV is less humanoid-specific, but its active robots, supply hours, fleet services, and software services are better current proof of robot fleet economics. RR stays close behind because it is the more direct public robot-platform and RaaS proxy, but the delayed March 2026 filing condition limits the current read.

TSLA Tesla, Inc.

Primary public humanoid platform proxy through Optimus, internal factories, autonomy data, AI infrastructure, and a large balance sheet.

Market cap$1.48T
Next earningsJul 22, 2026
Latest qtr revenue$22.387B

Role in stack

TSLA owns the potential finished humanoid platform loop: robot body roadmap, factories for internal trials, FSD and Robotaxi data, AI training infrastructure, service footprint, and a path to external customers if Optimus moves beyond internal use.

Revenue mix

Q1 2026 revenue was $22.387B. Q2 deliveries were 480,126 vehicles, Q2 storage deployments were 13.5 GWh, and the lane records 1.28M active FSD subscriptions in the Q1 update. Optimus revenue is not separately disclosed.

Proof burden

Factory task counts, paid robot units, intervention rate, uptime, unit cost, safety, service burden, internal productivity, and a path from AI capex to FCF. Strong auto or energy results do not by themselves prove humanoid economics.

SERV Serve Robotics Inc.

Best current public fleet-economics comparator through active robots, supply hours, fleet services, software services, and field support.

Market cap$488M
Next earningsNot confirmed
Latest qtr revenue$2.984M

Role in stack

SERV gives the page the clearest public read on robot fleet operations while its commercial base sits in delivery robots and healthcare automation: active robots, remote support, software services, customer concentration, and revenue per deployed capacity.

Revenue mix

Q1 2026 revenue was $2.984M, split between $1.958M fleet services and $1.026M software services. The lane records 812 daily active robots and 10,295 daily supply hours.

Proof burden

Revenue per robot, operator-to-robot leverage, remote intervention decline, service margin, customer retention, Diligent contribution, operating cash burn, and ATM discipline. Q1 gross loss was $9.001M and operating cash use was $41.422M.

RR Richtech Robotics Inc.

More direct robot-platform and RaaS proxy than SERV, with commercial robots, industrial robots, Dex optionality, and data-services ambition.

Market cap$438M
Next earningsNot confirmed
Latest qtr revenue$1.147M

Role in stack

RR sells and deploys commercial and industrial robots, is trying to grow RaaS and leasing/service revenue, and frames data services and Dex as future platform routes. It is the cleaner robot-OEM/RaaS concept, but still pre-scale.

Revenue mix

The latest filed local quarter in the lane was Q1 FY2026 with $1.147M revenue, including $319K of RaaS revenue and $1.722M of RaaS-related remaining performance obligations. SEC submissions still showed an NT 10-Q for the March 31, 2026 period as of the July 5 check.

Proof burden

Clean delayed filing, RaaS and RPO growth, customer revenue by product family, deployment density, service cost per robot, uptime, gross margin after support costs, data-services revenue, and share-count discipline.

PDYN Palladyne AI Corp.

Autonomy software and manufacturing transition monitor with platform relevance; current evidence sits in defense autonomy and production transition rather than finished humanoid deployment.

Market cap$276M
Next earningsNot confirmed
Latest qtr revenue$3.538M

Role in stack

PDYN contributes the autonomy-software and production-transition angle: SwarmOS, BRAIN, Palladyne Pilot, IQ, engineering services, and precision manufacturing. It fits platform learning and robot autonomy more than humanoid OEM ownership.

Revenue mix

Q1 2026 revenue was $3.538M. Backlog was $17.3M, total estimated contract value was $23.4M, and FY2026 revenue guidance was $24M-$27M. Software revenue is not yet separately visible at scale.

Proof burden

Backlog conversion, first-article approvals, named defense or commercial autonomy awards, software attach, gross-margin improvement, lower burn intensity, and limited ATM usage before the story becomes self-funding.

KSCP Knightscope, Inc.

Autonomous security robot and managed-security monitor with a real revenue step-up but the weakest funding and margin profile in the basket.

Market cap$33M
Next earningsNot confirmed
Latest qtr revenue$6.016M

Role in stack

KSCP leases autonomous security robots, sells emergency communication devices, operates monitoring software, and added human-security services through Event Risk. Its platform evidence is security automation and managed services, so humanoid relevance is only an adjacent robot-fleet monitor.

Revenue mix

Q1 2026 revenue was $6.016M, including $3.665M from Core Technology and $2.351M from Acquired Security Force. Consolidated gross margin was only $0.465M, and backlog was about $0.8M as of May 8.

Proof burden

Full-quarter KSF retention, Core Technology margin, ASR/ECD demand, K7 commercialization, operating cash burn reduction, cash runway, financing terms, and share-count control.

Basket source note: market caps use 2026-07-02 local closes times local weighted shares outstanding. Next earnings dates are left as not confirmed unless the filed local lane supplied a date. The right rail chart order follows the final tournament order: TSLA, SERV, RR, PDYN, KSCP.

Source-only monitors: Unitree, Figure, Agility Robotics, 1X, Apptronik, Boston Dynamics, UBTECH, Xiaomi, XPeng, Toyota, Honda, Hyundai, Doosan, and Kawasaki for direct or regional platforms; HSAI, FANUC, Yaskawa, Siemens, Keyence, Mitsubishi Electric, Daifuku, Omron, SMC, Nabtesco, and Harmonic Drive for regional ecosystem control points. On this page they stay in source-only monitor lists instead of ranked finished-platform cards.

What Confirms Or Weakens

Area What confirms What weakens or invalidates Watch next
01Platform thesisFinished robot control
What confirms

A company discloses finished robot hardware, control software, teleoperation, service tooling, and commercial deployment evidence in the same economic loop.

What weakens or invalidates

The evidence remains demos, parent-company optionality, supplier content, or broad automation language without paid robot units or task economics.

Watch next
  • Paid units
  • Task scope
  • Robot platform disclosure
  • Customer names
02Paid fleet economicsRevenue and margin
What confirms

Robot revenue, utilization, RaaS or lease revenue, service attach, gross margin, support cost, and customer retention improve together.

What weakens or invalidates

Pilots do not become repeat orders, robot count rises without revenue density, or gross loss and support labor scale with the fleet.

Watch next
  • Revenue per robot
  • Gross margin
  • Service cost
  • Repeat orders
03Fleet data and controlOperating proof
What confirms

Companies disclose active robots, supply hours, task success, intervention rate, uptime, incidents, software updates, or operator leverage.

What weakens or invalidates

Fleet logs stay private, human rescue remains high, incidents rise, or management only reports generalized AI and robotics milestones.

Watch next
  • Active robots
  • Supply hours
  • Intervention rate
  • Uptime
04Regional ecosystemSupply-chain split
What confirms

Named suppliers, manufacturing regions, actuator, sensor, control, compute, and service dependencies become visible through orders, customer pull, or policy support.

What weakens or invalidates

Regional monitors are only broad automation vendors, source-only OEMs lack filed economics, or export controls and standards split the stack without investable revenue.

Watch next
  • China deployments
  • U.S. AI-first pilots
  • Japan/Korea OEMs
  • Export controls
05Disclosure and fundingPer-share proof
What confirms

Timely filings, audited revenue detail, balance-sheet runway, non-dilutive customer funding, and segment disclosure show robot economics without hiding the burn.

What weakens or invalidates

RR filing delays persist, microcaps fund losses through equity, going-concern language worsens, or broad parent businesses hide the robotics signal.

Watch next
  • RR 10-Q status
  • ATM usage
  • Cash burn
  • Share count
06Data freshnessChart and source checks
What confirms

Local daily_ohlc stays current, report API charts render, and security lanes refresh after earnings, filings, funding updates, and deployment news.

What weakens or invalidates

API remains unavailable, lineage freshness lags, a new market session changes setup labels, or official filings update the stale local lane evidence.

Watch next
  • API status
  • Daily OHLC
  • SEC filings
  • Lane refresh

Source Trail

Routes used for this node
Route Use it for Primary links
Paper priorsFinished-platform role, OEM-level market framing, software and ecosystem lag, China deployment-led curve, Western AI-first curve, and first deployment use cases.Citrini humanoid primer; Roland Berger 2026 extraction; Roland Berger PDF
Ranked knowledge lanesCompany setup, proof burden, segment evidence, fleet KPIs, RaaS, backlog, cash burn, financing, and source gaps.TSLA; SERV; RR; PDYN; KSCP; Industrials; Information Technology
Regional monitorsSource-backed ecosystem control points and non-card regional context. These names map regional exposure; this page routes them as regional monitors instead of ranked finished-platform cards.HSAI; FANUC; Yaskawa; Siemens; Keyence; Hyundai, Toyota, Honda, Doosan, Kawasaki, Unitree, Figure, Agility, 1X, Apptronik, Boston Dynamics, UBTECH, Xiaomi, XPeng.
Official SEC checkRR filing freshness. The official submissions feed was checked on 2026-07-05 because the local lane flagged a delayed March 31, 2026 10-Q; the recent filing list still showed the 2026-05-15 NT 10-Q and no later matching 10-Q in the returned recent forms.SEC submissions for Richtech Robotics
Discovery and DuckDB methodsRead-only local coverage, market-cap calculation, and weekly setup labels. TSLA, SERV, RR, PDYN, and KSCP have daily_ohlc rows through 2026-07-02.node metadata; discovery status TICKER --json; discovery lineage-status --ticker TICKER --dataset daily_ohlc --limit 3 --json; read-only DuckDB on daily_ohlc and instruments.
API chart routeRight-rail chart rendering. Each rank button and metrics strip calls the selected-security panel through data-chart-ticker. Weekly charts should derive from daily rows with a three-year visible horizon, 20-week EMA, 100-week EMA, and volume subgraph./api/securities/TICKER/chart?frequency=weekly&window=3y&as_of=latest; no embedded OHLC arrays; no static horizontal chart levels. Local API check on 2026-07-05 returned connection refused at 127.0.0.1:8765.
Ranking processTen xhigh primary research agents and one xhigh merge selector resolved the basket. The merge kept TSLA first, moved SERV above RR for current fleet evidence, retained RR as the more direct RaaS/platform proxy with filing caveat, and kept PDYN and KSCP as lower-conviction monitors.Final order: TSLA, SERV, RR, PDYN, KSCP.
Data limitationsLocal discovery supports price, volume, chart routing, and computed market-cap context. It does not prove humanoid revenue. Papers are priors; filed lanes, official SEC status, and read-only discovery checks are the posterior.Refresh if a new filing, earnings release, customer contract, robot deployment disclosure, financing update, or local API behavior changes.

Categories