The DUAL Autonomy Endorsement
Powered by MovementAI
A proposed white-label initiative for DUAL and the Howden Group to evaluate a patent-pending pre-loss risk-intelligence framework that detects deteriorating operating conditions, quantifies risk velocity, and supports governed intervention before loss occurs.
Standardizing Risk in the Built Environment
Discussion draft: projections and proposed product architecture. All scores, pricing bands, premium effects, valuation effects, loss scenarios, adoption figures, timelines, staffing levels, and commercial terms in this proposal are forward-looking hypotheses for joint validation. They are not historical results, actuarial indications, insurance quotes, coverage commitments, legal conclusions, or guarantees. Final terms remain subject to data diligence, actuarial review, underwriting approval, regulatory analysis, security review, legal documentation, and mutually executed agreements.
Executive Summary
Movement intelligence is the missing pre-loss layer between connected buildings and autonomous operations.
Buildings cannot become safely autonomous merely by adding sensors, robots, or software. They must first understand how people, machines, materials, schedules, spaces, controls, and external conditions interact, and recognize when those interactions are converging toward loss. MovementAI proposes that intelligence layer for DUAL and Howden.
Immediate ask: authorize a jointly scoped 90-day feasibility sprint to assess data availability, retrospective signal quality, governance requirements, and the case for a prospective pilot.
See risk forming
MovementAI converts fragmented physical-operating signals into leading indicators of control degradation, unsafe coupling, cascade proximity, and risk velocity, before those conditions mature into loss.
Intervene with authority
Qualified recommendations include the evidence, confidence, expected propagation path, intervention horizon, and accountable decision-maker. High-consequence actions remain human-governed.
Prove what happened
Every signal, model version, recommendation, decision, execution record, and outcome forms a traceable evidence chain for underwriting, risk engineering, claims review, and portfolio learning.
The AutonomousReady Path
Autonomy is therefore an earned operating state, not a technology label. Each level requires stronger data quality, tested controls, explicit authority, fail-safe behavior, and evidence that the prior level performs as intended.
Underwriting
A dynamic view of current control integrity, not only historical loss and periodic attestations.
Risk engineering
Earlier, prioritized intervention while preventable operating conditions can still be changed.
Claims
Time-sequenced evidence separating source integrity, decisions, execution, outcome, and causation review.
Portfolio strategy
A consistent framework for comparing readiness, intervention performance, and emerging concentration risk.
The strategic proposition is subject to technical, actuarial, legal, security, regulatory, and operational validation. MovementAI augments professional judgment; it does not bind coverage, determine compliance, establish causation, or independently execute high-consequence actions.
The Precedent
You did it with Cybeta for Cyber.
We are the Cybeta for the Physical World.
In 2026, Howden announced its acquisition of Cybeta's intellectual-property assets to support faster, more precise cyber-risk insight across underwriting and placement. That public precedent suggests a relevant strategic pattern, not proof of fit, for evaluating a comparable data-and-analytics layer for physical operations.
We are not pitching "Howden uses our score." We are pitching: DUAL builds a new product category around our score, and Howden distributes it globally.
Joint ambition: determine whether governed leading indicators of operational instability can provide underwriters with reproducible, decision-useful evidence before those conditions mature into incidents, claims, or systemic loss.
Proposed IP Differentiation
MovementAI describes two patent-pending applications, MOMENTUM-001PROV and MOMENTUM-002PROV, as containing more than 650 combined claims. Application status, ownership, scope, validity, enforceability, freedom to operate, and competitive effect require documentary diligence and patent-counsel review. The proposed commercial distinction is an orchestration layer that scores physical movement risk across the asset lifecycle.
- Trade-Species Classification: A proposed framework for assigning constraints and dependencies to human and machine movement.
- Mathematical Risk Forecasting: Patent-pending methods intended to identify dependency stress and potential schedule cascades before loss.
- Proposed Defensible Scope: Subject to diligence, the application portfolio may support differentiated orchestration and scoring capabilities; no exclusivity or market outcome is guaranteed.
The Market Reality
Why Hardware Integration Has Stalled
Traditional controls document incidents and manage exceptions after risk has materialized. The proposed framework is designed to identify precursor conditions, estimate escalation potential, and support accountable intervention while loss remains preventable.
Integration & Safety Liability
A 2025 Frontiers scoping review documents safety risks associated with robotics and automation in construction, while broader literature identifies human–robot collaboration, technical integration, workforce readiness, privacy, legal concerns, and uncertain economics as adoption barriers. The evidence supports a multi-factor integration problem; it does not establish liability as the sole or primary cause of stalled adoption.
Cyber-Physical Risk Exposure
Connected operational technology can translate digital compromise into safety, availability, environmental, and physical-loss consequences. Gartner's earlier executive-liability forecast was tied to 2024 and should be treated as historical market context, not a current prediction or measured outcome. Its 2024 research continues to highlight unmanaged legacy and cyber-physical systems as a zero-trust gap.
The Orchestration Solution
Endpoint manufacturers, building systems, construction platforms, security tools, and insurers each address part of the problem. MovementAI proposes a cross-system pre-loss intelligence layer that converts fragmented operating signals into time-bounded risk states, escalation pathways, and governed intervention recommendations. Validation must establish whether those signals are reliable, incremental, and materially earlier than existing controls.
The DUAL Underwriter Interface
From Retrospective Evidence to Governed Pre-Loss Intelligence
Conventional underwriting relies heavily on historical loss experience, periodic surveys, and static control attestations. The proposed AutonomyScore™ feed would add traceable leading indicators: current control integrity, emerging operational coupling, risk velocity, confidence, and the time remaining for authorized intervention.
- APEX Readiness Projection: Pre-underwrite policies during construction.
- Safety Evidence: Traceable workflow records subject to source and human verification.
- Cascade Proximity: A leading indicator of how rapidly connected operating constraints are converging toward a loss-producing state.
Overall AutonomyScore™
94/100
APEX Readiness: Level 4
Partnership Structure
The Four-Tier Architecture
A proposed operating model for diligence, from data ingestion through potential product and reinsurance evaluation. Each tier is conditional and subject to separate approval.
Tier 1
Data & Scoring Layer
MovementAI would provide governed risk signals through an API into DUAL's underwriting workflow. The proposed evidence set combines cascade proximity, control integrity, crisis resilience, data confidence, and source lineage. These signals augment, not replace, professional underwriting and are intended to reveal material changes in exposure earlier than periodic reviews.
Tier 2
DUAL Product Creation
Subject to validation and approval, DUAL could develop an Autonomous Operations Endorsement or underwriting program using agreed score bands for referral and control review. DUAL would retain authority over eligibility, policy language, pricing, and coverage terms.
Tier 3
Howden Distribution
If the product is validated and launched, Howden could evaluate distribution through its global network, which reports 24,000 people. Eligible clients, territories, industries, enrollment, and broker workflows would be defined through a jointly approved market plan.
Tier 4
Reinsurance (Howden Re)
If evidence supports the framework, Howden Re could evaluate whether governed score bands add useful information to treaty analysis. Reinsurance use, pricing influence, and market adoption would remain independent future decisions, not assumed outcomes.
Proposed Insurance Blueprint
From Score Concept to Governable Product
These are design requirements for joint development, not existing policy terms or a commitment to provide coverage.
Proposed form
A jointly designed risk-control endorsement or underwriting program centered on continuous pre-loss intelligence for assets using autonomous or cyber-physical systems. Whether it is admitted, E&S, delegated, manuscript, or risk-engineering-led remains an open legal and regulatory workstream.
Risk appetite
Define eligible occupancies, construction types, territories, total insured values, technology classes, autonomy levels, minimum controls, minimum data history, and prohibited or referral risks before any quote is contemplated.
Coverage architecture
Product counsel and underwriters would define covered causes of loss, exclusions, sublimits, deductibles, warranties, control requirements, score-change consequences, cancellation rights, and whether any premium modification is permitted.
Claims protocol
Preserve the complete risk-state history: precursor signals, escalation pathway, intervention window, recommendations, human decisions, control execution, timestamped source records, and model version. Give authorized adjusters controlled evidence access, separate data integrity from causation, and maintain a documented challenge and correction process.
Operating role
MovementAI is proposed as a technology and risk-intelligence provider, not an insurer, broker, adjuster, regulator, certification authority, or final underwriting decision-maker. DUAL retains underwriting and coverage authority.
Decision rights
The score informs, not replaces, professional judgment. Underwriters can refer, override, or decline with a recorded rationale; insureds can challenge material data errors; emergency and life-safety authority always supersedes automated recommendations.
Hypothetical Case Studies
From Reactive Incident Management to Predictive Pre-Loss Orchestration
Illustrative scenarios showing how precursor-state detection and human-governed intervention are intended to interrupt operational instability before loss. They are not actual deployments, claims, or validated loss outcomes.
Case 1 · Occupied renovation
Stopping a water-loss cascade before pressure testing
Reactive model
A high-pressure plumbing test proceeds while sensitive equipment is being installed one floor below. Conventional controls identify the conflict only after a leak alarm, damaged equipment, and tenant disruption trigger an incident response.
Precursor state
Permit timing, contractor access, pressure-test schedule, vertical-zone adjacency, tenant move-in activity, and the status of isolation controls begin converging into a high-proximity risk state.
Governed intervention
MovementAI qualifies the overlapping movements, identifies the missing isolation verification, estimates a narrowing intervention horizon, and recommends a temporary hold. The authorized site lead verifies the condition and reschedules the test under existing permit authority.
Hypothetical pre-loss result
The incompatible activities are separated before the test begins. The evidence chain preserves the source signals, recommendation, approval, revised schedule, and completed control check for later risk-engineering and underwriting review.
Case 2 · Distribution and logistics
Preventing dock congestion from becoming business interruption
Reactive model
A late inbound trailer collides with a compressed outbound schedule, blocked staging space, and a forklift charging constraint. Teams respond after queues form, labor idles, temperature-sensitive goods wait, and service commitments are threatened.
Precursor state
Carrier ETA drift, dock occupancy, staging capacity, equipment availability, labor windows, and cold-chain dwell limits show accelerating dependency stress before the yard reaches gridlock.
Governed intervention
MovementAI detects rising risk velocity, qualifies alternate dock and staging paths, and recommends resequencing two arrivals while protecting the temperature-controlled load. A dock supervisor reviews and authorizes the revised flow.
Hypothetical pre-loss result
The bottleneck is relieved while options remain available, reducing the likelihood of spoilage, collision exposure, and prolonged interruption. The operational thread records what changed, why it changed, who approved it, and whether dwell time returned to tolerance.
Case 3 · Autonomous building operations
Containing control degradation before an unsafe machine movement
Reactive model
A service robot receives a routine route command while an access-control feed is stale and an occupied work zone has changed. A conventional workflow discovers the mismatch only after a near miss, emergency stop, or manual complaint.
Precursor state
Credential freshness, route clearance, endpoint readiness, work-zone status, network health, and machine identity confidence fall out of alignment, lowering control integrity before the mission starts.
Governed intervention
MovementAI rejects automatic qualification, places the mission in a reversible safe-hold state, explains the conflicting evidence, and requests human verification. The facilities operator updates the route and releases the mission only after controls are restored.
Hypothetical pre-loss result
The machine does not enter the uncertain zone. The preserved record links the stale input, qualification failure, safe hold, operator decision, corrected route, and eventual outcome, supporting governance without claiming that automation risk has been eliminated.
The Legal Architecture
Proposed Technical Differentiation for Underwriters
The orchestration and evidence architecture is presented for technical, patent, security, and underwriting diligence. Pending claims may support differentiation, but do not guarantee validity, coverage, exclusivity, or commercial outcomes.
Systemic Risk Mitigation
Construction Cascade Proximity Scoring (CCPS) is proposed as a forward-looking measure of dependency stress: the degree, velocity, and confidence with which connected constraints are converging. Rather than reporting a failure after it occurs, CCPS is designed to surface the precursor state, intervention window, and principal drivers before propagation reaches a loss threshold.
Machine Trust & Cyber-Physical Security
The proposed Machine-to-Machine Trust Validation control would check available device credentials and cryptographic signatures before eligible commands are released. Its effectiveness depends on identity assurance, key security, integration quality, authorized human control, and testing; it does not eliminate cyber-physical liability.
Verified Operational Evidence
The platform is proposed to assemble timestamped safety-workflow evidence and operational logs from available source systems. It can support authorized safety personnel, but does not itself determine OSHA compliance, replace a competent person, issue regulated permits, or prove the truth or completeness of every source event.
Cross-Portfolio Risk Normalization
Federated Learning and Portfolio Benchmarking could support privacy-aware comparison of control performance across participating assets. Any model update would require approved governance, validation, and change control; improvement at one asset would not automatically alter another asset's risk profile.
Degraded Autonomy & Fallback Protocols
The proposed reliability design uses predefined Safe-Hold States, referral, and human override when network, compute, or source systems fail. Production effectiveness would depend on tested integrations, local controls, and approved failure playbooks.
Crisis & Climate Resilience
NOAA reports 27 U.S. billion-dollar weather and climate disaster events in 2024. Against that catastrophe-risk backdrop, the proposed system would ingest approved climate and disaster forecast data to inform human-governed prioritization, safe-hold, and rerouting decisions; effectiveness remains subject to integration and scenario testing.
Occupied Renovation Liability
The proposed Concurrent Renovation Thread Protocol would identify conflicting worker, tenant, equipment, and schedule movements in occupied assets so authorized teams can intervene before conditions escalate. It is intended to reduce exposure, not guarantee claim prevention.
Regulatory Compliance Shield
A governed evidence record could support Digital Product Passport and audit workflows by linking approved source events, decisions, controls, and outcomes. Regulators determine acceptability, and the record does not itself establish compliance or eliminate audit exposure.
The Incumbent Defense
Building-management, construction, safety, and workflow platforms each address part of the operating environment. MovementAI's proposed distinction is a cross-system qualification layer for the live interaction of people, machines, materials, spaces, controls, and schedules, subject to competitive and patent diligence.
Frictionless Deployment & Privacy
Initial feasibility analysis can begin API-first using available owner systems, without requiring new hardware. Higher-confidence real-time controls may require additional sensors or integrations. Tokenization and minimization can reduce privacy exposure, but GDPR, CCPA, employment, biometric, and local requirements must be assessed for each deployment.
Full Lifecycle Capture
The "Builder's Risk to Permanent" Bridge
The highest risk phase of any asset is ground-up construction. Through MovementAI Build™, we accumulate an 18–24 month Construction-Phase Operational Thread Record (CPOTR) from groundbreaking to completion.
The proposed pilot would test whether changes in cascade proximity, control integrity, and intervention response provide earlier and more discriminating evidence for Builder's Risk than schedule variance and incident reporting alone. Any premium effect requires actuarial support and underwriting approval. “Born Certified” is a proposed readiness designation, not a regulatory, insurance, engineering, or safety certification.
The building's operational AI starts with a calibrated baseline rather than "zero history," establishing a permanent structural advantage in operational accuracy.
Furthermore, underwriters can utilize our APEX Readiness Projection. By analyzing infrastructure choices (sensors, compute) during construction, DUAL can mathematically pre-underwrite the asset's permanent P&C policy long before the building even opens.
Deployment & Reliability
Capability Must Match Data and Control Maturity
Initial analysis may require no new hardware; real-time physical orchestration requires trustworthy inputs, tested integrations, explicit authority, and resilient operating controls.
Level 1, Desktop feasibility
Use schedules, policies, asset registers, incident history, claims taxonomy, and exported operational records to test data mapping and retrospective score logic. No real-time control or loss-prevention claim is available at this level.
Level 2, Connected observation
Integrate available access control, work-order, BMS, IoT, weather, logistics, and safety systems through approved APIs or secure files. Convert time-sequenced observations into leading indicators with source lineage, signal latency, confidence, missing-data flags, risk velocity, and estimated intervention horizon.
Level 3, Human-governed intervention
Route qualified pre-loss recommendations to authorized operators while a defined intervention window remains open. Record the detected precursor state, expected propagation pathway, acknowledgement, decision, reason, execution evidence, and outcome; do not independently execute high-consequence actions.
Level 4, Approved automation
Only narrowly defined, tested, reversible actions may be automated after hazard analysis, legal approval, cybersecurity testing, owner authorization, fail-safe design, emergency override, and documented accountability.
Production service profile
Before launch, agree supported source systems, refresh rates, data quality thresholds, uptime, latency, support, maintenance, incident notification, recovery objectives, degraded mode, data reconciliation, and responsibility for third-party outages.
Failure handling
Low confidence, stale or conflicting data, source failure, suspected spoofing, model drift, or unsafe conditions trigger referral, safe hold, human review, or suspension according to approved playbooks. Every material exception remains auditable.
Actuarial Validation
A Credible Evidence Path, not a 90-Day Proof Claim
The first 90 days establish feasibility and leading indicators. Loss-cost and pricing conclusions require larger populations, historical data, controls, and time.
1. Feasibility
Inventory available exposure, claims, operational, safety, schedule, and control data; document lineage and missingness; establish lawful access and a common claims taxonomy. Exit only if data can support defensible analysis.
2. Retrospective study
Apply frozen score logic to historical periods; compare score components with frequency, severity, near misses, delays, and control failures; use matched cohorts and normalization for occupancy, geography, value, construction type, and catastrophe exposure.
3. Prospective pilot
Operate in shadow mode and test whether precursor signals consistently surface before conventional alerts. Measure intervention lead time, risk-state stability, calibration, false-positive and false-negative rates, operator response, prevented propagation, downtime, near misses, and evidence completeness. A short pilot tests operations, it does not establish long-term loss-ratio improvement.
4. Actuarial translation
Estimate indicated effects only where credible; report uncertainty and sensitivity; separate correlation from causation; require actuarial sign-off before translating score bands into referrals, credits, surcharges, expected-loss assumptions, or pricing.
5. Independent challenge
Use an agreed reviewer to test methodology, bias, leakage, stability, reproducibility, and adverse selection. Publish limitations and unresolved exceptions alongside favorable findings.
6. Multi-year monitoring
Track fire, water, injury, cyber-physical, property, and business-interruption outcomes on their appropriate development timelines; recalibrate or suspend use if drift, poor performance, or unintended impacts exceed approved tolerances.
Governance & Accountability
Controls for a High-Consequence Decision System
The design target is explainable, challengeable, human-governed intelligence aligned with the NAIC AI Model Bulletin and NIST AI RMF, not autonomous insurance decision-making.
Model governance
Maintain score definitions, features, weights, confidence, approved uses, prohibited uses, validation status, version history, change approvals, drift thresholds, overrides, appeals, and rollback procedures aligned to insurer governance expectations.
Human authority
No automated recommendation should independently bind coverage, admit liability, issue a regulated permit, disable emergency egress, or initiate a high-consequence physical action. Authorized personnel approve interventions, with emergency override and safe-hold controls.
Data and privacy
Contractually define controller/processor roles, purpose limitation, lawful basis, minimization, tokenization, retention, deletion, access rights, subprocessors, cross-border transfer, breach notification, and impact assessments. Compliance is assessed, not guaranteed by design language.
Security and resilience
Agree encryption, identity, least privilege, key management, logging, vulnerability management, testing, incident response, data residency, supplier controls, uptime, latency, support, recovery-time and recovery-point objectives before production.
Input integrity
Cryptography can demonstrate record integrity after capture; it cannot prove that a sensor, person, or source system reported truthfully. Source assurance, calibration, reconciliation, confidence scoring, anomaly detection, and evidence provenance remain required.
Contractual risk
Definitive agreements should address IP and derived-data ownership, confidentiality, audit rights, service levels, professional and cyber insurance, indemnities, liability caps, consequential loss, regulatory cooperation, termination, data return, and transition support.
Go To Market
The Initial 90-Day Feasibility Sprint
A proposed discovery and operating-validation phase. Public launch, branding, pricing, and coverage would occur only after agreed evidence, governance, legal, regulatory, and commercial gates are satisfied.
Feasibility, Retrospective Analysis & Operating Pilot
- 1Asset Selection & The Belpointe Fast-TrackDUAL and MovementAI would nominate an initial feasibility cohort, subject to owner consent and data availability. Belpointe assets may be evaluated as proposed reference candidates only after deployment status, permissions, history, and suitability are documented; this proposal does not assert completed “Born-Certified” deployments.
- 2Score GenerationMovementAI maps approved source data into a baseline risk state, then evaluates whether changes in precursor signals can be detected with sufficient lead time, confidence, stability, and explanatory evidence to support authorized action.
- 3The Actuarial Data BridgeDuring the sprint, the parties test whether score components can be reproducibly mapped to de-identified exposure and claim taxonomies. Any relationship is reported with uncertainty and limitations. The work does not “prove” a premium offset; actuarial translation follows only after credible retrospective and prospective evidence.
- 4Dedicated Resource AllocationProposed planning assumption: MovementAI supplies engineering and data-science leads; DUAL/Howden considers actuarial, underwriting, product, claims, legal, security, and regulatory representatives. Final staffing, effort, budget, data access, deliverables, and ownership are agreed in a signed pilot statement of work.
Proposed Post-Validation Market Education
If validation and approvals succeed, the parties could develop a counsel-reviewed market-education program. Any “first,” “standard,” performance, certification, or exclusivity claim would require substantiation and mutual written approval before publication.
- Co-authored White Papers on "Insuring the Autonomous Asset"
- Dedicated "Howden AutonomyScore" landing pages and branding
- Press tours focusing on DUAL's innovative new product line
- MovementAI stays in the background as "Powered By"
Commercial & Partnership Terms
A Negotiation Framework for Building the Category Together
Every commercial number remains a projection to be jointly modeled; this proposal creates no exclusivity, investment right, premium share, or binding obligation.
Discovery economics
The parties would agree a fixed discovery budget covering integration mapping, data preparation, actuarial analysis, security review, legal work, and independent validation. No fee or resource commitment is assumed until documented.
Projected production economics
Potential structures, per-score fee, annual platform license, implementation fee, or compliant revenue participation, are alternatives for diligence. Any GWP-linked compensation requires legal and regulatory review and is not presumed permissible.
Market sizing
A jointly approved business case should quantify eligible assets and premium by territory, adoption rate, average fee, implementation capacity, sales cycle, churn, support cost, and downside/base/upside cases. No market-size number is asserted in this draft.
Exclusivity
Any proposed exclusivity must define product, channel, territory, term, minimum performance, development milestones, reserved markets, IP rights, termination, change of control, and post-termination transition. Exclusivity is not granted by this proposal.
Brand and ownership
Working convention: DUAL would own underwriting decisions and policy language; MovementAI would retain its pre-existing platform, models, and IP; jointly created branding, improvements, derived data, and customer rights require negotiated allocation.
Scale and exit gates
Launch requires approved product language, regulatory pathway, validated use case, security acceptance, integration readiness, broker and underwriter training, claims procedures, support model, and economics. Failure criteria trigger remediation, scope reduction, suspension, or orderly exit.
Executive Reference
Glossary of Movement-Intelligence Terms
Plain-language definitions for the proposed technical and insurance framework. Terms describe decision-support concepts unless expressly validated and adopted in a final product.
- APEX Readiness Projection
- A MovementAI automation-readiness framework that evaluates whether an asset’s routes, access, charging, controls, compute, endpoints, and operating rules can support proposed autonomous missions. It identifies gaps and improvement priorities; it is not an engineering or insurance certification.
- AutonomousReady
- An evidence-based maturity state in which an asset can coordinate defined autonomous missions within tested control limits, human authority, fail-safe procedures, and auditable operating rules. It does not mean unrestricted or unsupervised autonomy.
- BORN / Born-Certified™
- A Momentum Labs evidence designation for a building that begins MovementAI during construction and enters occupancy with site memory, construction history, readiness baselines, and operational evidence already established. It is not a regulatory, safety, engineering, or insurance certification.
- Cascade Proximity
- The degree to which interconnected constraints, such as schedule, access, capacity, equipment, control, or environmental conditions, are converging toward a loss-producing state. It describes nearness and escalation pressure, not a guaranteed event.
- CCPS
- Construction Cascade Proximity Scoring: the proposed method for evaluating dependency stress, convergence, risk velocity, confidence, and intervention lead time during construction or renovation.
- Control Integrity
- The current evidence that a preventive or mitigating control is present, authorized, functioning, timely, and appropriate for the operating condition it is expected to govern.
- CPOTR
- Construction-Phase Operational Thread Record: the time-sequenced history of movements, constraints, interventions, exceptions, controls, and outcomes accumulated from construction through handover.
- Evidence Chain
- The linked record of source signals, lineage, model version, qualification logic, recommendation, human decision, execution evidence, exception, and outcome used for review and challenge.
- Human-Governed Intervention
- A recommended action that an identified, authorized person must review, approve, modify, or reject. Emergency and life-safety authority always supersedes the recommendation.
- Intervention Horizon
- The estimated time remaining in which an authorized action could materially reduce propagation or severity before the risk state crosses a defined threshold.
- Movement Friction
- Any condition that delays, blocks, reroutes, conflicts with, or degrades the intended movement of people, goods, machines, services, energy, or information through an asset.
- Movement Intelligence
- The continuous interpretation of physical-operating signals and dependencies to identify precursor states, qualify risk, recommend governed action, and preserve outcome evidence before loss occurs.
- Movement Qualification
- The process of testing a proposed movement against readiness, access, policy, capacity, endpoint, safety, trust, and timing requirements before it proceeds.
- Precursor State
- An observable combination of conditions that can arise before an incident, such as control degradation, unsafe overlap, capacity compression, or accelerating dependency stress.
- Risk Velocity
- The rate and direction at which a qualified risk state is changing over time. Rising velocity can shorten the intervention horizon even when current severity remains moderate.
- Safe-Hold State
- A predefined, reversible operating condition used when confidence is low, signals conflict, a source fails, or proceeding could create unacceptable risk. It preserves human override and emergency egress.
- Signal Confidence
- A bounded assessment of whether available evidence is sufficiently complete, fresh, consistent, and trustworthy for the stated decision. Low confidence triggers referral or safe hold rather than hidden certainty.
- Operational Thread
- The continuous, time-sequenced connection between what the asset observed, what it recommended, what people decided, what was executed, and what outcome followed.
Evidence Appendix
Sources, Boundaries, and Diligence Status
External sources establish market context only. They do not validate MovementAI performance, pricing, coverage, or loss-reduction projections; those require the joint validation program above.
- 01
- 02
Howden, About Us
Howden reports 24,000 people and $51bn of client premiums - 03
- 04
- 05
- 06
- 07
- 08
Patent counts, application status, ownership, asset references, deployment status, score methodology, security posture, market size, financial assumptions, and all proposed economics remain subject to documentary diligence and counsel review before external reliance.