PROJECT MANAGEMENT
Can AI maintain coherent project knowledge across homeowner, architects, engineers, authorities, finance, contractors and suppliers?
GM30 / AI case study / Hypothesis
A living prototype within the SWIP ecosystem.
One house. One complete transformation. One continuous AI-supported project conversation.
Specific project experiment
PROPOSED / NOT YET MEASURED
Lower-floor excavation coordination.
Can AI-supported project management help the homeowner coordinate architect, RIB, geotechnics, building physics, acoustics, HVAC, electrical, cost planning, contractor and authorities around one complex excavation and lower-floor expansion?
AI may organise questions, records, comparisons and decisions. It does not establish feasibility, design structure, approve work or replace accountable professionals.
Coordination participants
ArchitectRIBGeotechnical adviserBuilding-physics adviserAcoustic consultantHVAC adviserElectrical adviserCost plannerContractorAuthorities
Evidence fields
Questions identified · Not yet measuredConsultants required · Not yet measuredDecisions · Not yet measuredDesign changes · Not yet measuredCost changes · Not yet measuredRisk identified early · Not yet measuredTime spent · Not yet measuredAI assistance · Not yet measuredProfessional verification · Not yet measured
Proposed project stages
01
Existing terrain
02
Excavation
03
Foundation-ready
04
Retaining walls + slab
05
Weather-tight shell
06
Technical installation
07
Interior fit-out
08
Studio / Gym / Cinema complete
The core hypothesis
Can AI become a coordination layer across the complete lifecycle of a home?
GM30 tests this question through one real residential transformation—from redesign and finance through engineering, construction and sustainable operation. It does not claim the result before evidence exists.
EFFICIENCY
Can AI reduce duplicated work, information-search time, administrative friction, unnecessary iterations and delays?
QUALITY
Can AI improve coordination, traceability, specification control, risk identification and decision continuity?
SUSTAINABLE LIVING
Can project intelligence remain useful after completion for energy, indoor climate, maintenance, materials and future upgrades?
FINANCING + PARTNERSHIPS
Can AI-assisted project management improve the conditions for external financing and partner participation—and help enable completion before 2030?
Market context
The market is already moving.
GM30 is not developing its AI hypothesis in isolation. Norwegian architecture and property-development actors are already beginning to commercialise AI-assisted workflows for tasks such as property and site analysis combined with professional architectural assessment.
01
PHASE 1
AI helps analyse development potential.
02
PHASE 2
AI assists design and option development.
GM30 hypothesis
AI becomes a continuous coordination layer across the entire transformation lifecycle.
ANALYSE→DESIGN→APPROVAL→FINANCE→ENGINEER→PROCURE→BUILD→DOCUMENT→COMMISSION→OPERATE→MAINTAIN
AI ASSISTS.
PROFESSIONALS DECIDE.
What the hypothesis tests
PROJECT MANAGEMENTEFFICIENCYQUALITYFINANCING READINESSSUSTAINABILITYTRACEABILITYLONG-TERM BUILDING PERFORMANCE
Positioned as a standard GM30 is testing the potential for—not an achieved industry standard.
Research dimension 05
Hypothesis / Not yet proven
Financing + partnerships.
Can AI-assisted project management increase the ability of a privately owned residential transformation project to attract external financing, strategic partnerships, supplier contributions and sponsorship—and help enable completion of GM30 before 2030?
This is a hypothesis to be measured. AI does not obtain financing, and no contribution or completion outcome is implied.
- 01AI-ASSISTED PROJECT MANAGEMENT
- 02BETTER PROJECT STRUCTURE
- 03BETTER DOCUMENTATION
- 04CLEARER BUDGET + RISK
- 05FASTER DUE DILIGENCE
- 06STRONGER PARTNER PROPOSITION
- 07GREATER EXTERNAL CONFIDENCE
- 08POTENTIAL FINANCING / PARTNER CONTRIBUTION
- 09GM30 COMPLETION BEFORE 2030
01
DEBT FINANCING
Banks / construction finance / potential long-term financing
02
PUBLIC SUPPORT
Eligible energy / housing-quality support
03
STRATEGIC PARTNER CONTRIBUTION
Products / materials / technology / professional services / discounts
04
SPONSORSHIP / COMMERCIAL CONTRIBUTION
Cash or in-kind support under a defined commercial partnership
Categories remain separate in measurement and reporting.
Wider research question
Can an AI-assisted owner-side project manager give an individual homeowner some of the coordination capability normally associated with a professional property-development organisation?
AI may assist with
Partner identificationResearchProject briefsFinancial scenario preparationDocumentationDue-diligence preparationMeeting preparationFollow-upComparison of proposalsPartner pipeline managementDecision loggingProject knowledge
AI does not make credit decisions.
AI does not negotiate authority away from the owner.
AI does not replace legal, financial, engineering or construction professionals.
Measurement framework
Evidence, including failure.
The eventual case study should report successful, declined and inconclusive outcomes. If AI-assisted project management does not materially improve results, that remains a valid finding.
Qualified targetsTargets contactedMeetingsResponse rateProposalsFinancing indicationsConfirmed financingCash contributionSupplier discount valueProduct contribution valueProfessional-hours contributionTotal external project valueTime to decisionDocuments requestedDocuments suppliedReason acceptedReason declinedAI time savedHuman verification required
Important outcome metric
Target completion: before 2030
Hypothesis / Not yet proven
The continuous project conversation
Information that can remain connected.
The experiment asks whether context can travel from the earliest idea through the operational life of the house—without replacing professional judgement.
- 01
IDEA
- 02
EXISTING BUILDING
- 03
DESIGN
- 04
REGULATION
- 05
FINANCE
- 06
ENGINEERING
- 07
PROCUREMENT
- 08
CONSTRUCTION
- 09
COMMISSIONING
- 10
OPERATION
- 11
MAINTENANCE
- 12
FUTURE UPGRADE
AI coordinates information.
Professionals remain accountable.
From fragmentation to coordination
An opportunity to connect the record.
The industry already works through many necessary formats and specialist systems. GM30 explores whether an AI-supported layer can make their relationships easier to follow.
Today / Useful records, often separate
Emails
PDFs
Drawings
Spreadsheets
Portals
Quotes
Product sheets
Meeting notes
Individual knowledge
Separate systems
GM30 / Coordination hypothesis
Shared project knowledge
Structured decisions
Traceable documentation
Connected materials
Connected costs
Connected drawings
Connected performance
Continuous history
SWIP ecosystem
The wider context. Not the primary identity.
SWIP explores coordination of people, services, activities and resources. GM30 tests that philosophy at the scale of architecture, professionals, finance, materials, construction, technology and building performance.
Future knowledge layer
Connected records, searched as one conversation.
This is prepared as a future architecture only. Each connection must be supported by a real source; an empty field remains unknown.
PERSONPROJECT ROLEROOMDRAWINGDECISIONREQUIREMENTREGULATIONMATERIALPRODUCTSUPPLIERQUOTECOSTCONTRACTCHANGEINSTALLATIONEPDENERGY IMPACTMAINTENANCEPERFORMANCELESSON LEARNED
Ask GM30 / Future supported questions
Answers must follow evidence.
Why was this decision made?
Which drawing approved this change?
Which engineer reviewed it?
What did the original specification require?
Which product was ultimately installed?
What did it cost?
Does it have an EPD?
What maintenance does it require?
How does it affect energy use?
What changed during construction?
Is the measured result consistent with the design assumption?
What should be considered if this part of the house is changed in ten years?
Industry-standard hypothesis / Not a claim
Can a continuous AI-supported project record become a standard way of coordinating residential redesign, rebuilding and sustainable operation?
GM30 explores whether this model could inform a future coordination standard. It tests the potential and may provide learnings relevant to other projects; it has not established a standard.
REAL HOMEREAL PROFESSIONALSREAL BUILDAI-SUPPORTED COORDINATIONMEASURED RESULTSREUSABLE LEARNING
GM30 is not a demonstration of AI technology. It is a test of a possible new way of coordinating one real building project.