Skip to content
Construction Metrics

FeatureAI

AIWhite paperSuffolk and MITPublished September 16, 2026

Suffolk and MIT model 17% to 20% AI cost savings on one multifamily project

Suffolk’s model takes $32 million off a $179 million baseline, and 11 months off a 51-month timeline, for one completed multifamily project in San Francisco. The authors call the evidence “suggestive rather than definitive,” and no project has measured the savings.

Modeled cost cut
$32Mof a $179 million baseline, 18%
Modeled schedule cut
11 mo.of a 51-month timeline, 22%
Headline ranges
17% to 20% cost
22% to 25% schedule

Modeled estimateNot measured on any project

Construction holds $28 million of the $32 million the model takes out

ACQUISITION$27M baselineNo modeled savingsENTITLEMENT$1.4M, $0.05M cutDESIGN$6M, $1.2M cutPRECON$3M, $0.3M cutCONSTRUCTION$128M baselineMODELED CUT$28M22% of the phaseFINANCE$7M, $1.2M cutESCALATION$7M, $1.2M cut ACQUISITION$27M baselineno savingsENTITLEMENT$1.4M baseline$0.05M cutDESIGN$6M baseline$1.2M cutPRECONSTRUCTION$3M baseline$0.3M cutCONSTRUCTION$128M baseline$28M cutFINANCE$7M baseline$1.2M cutCOST ESCALATION$7M baseline$1.2M cut
Baseline cost of the sample project by phase, $ million, with the modeled AI savings in red at the end of each phase. Suffolk’s model of a 180,000-square-foot multifamily development finished in 2024 (Table 1). Phases add to the $179 million total after rounding. Source: Suffolk, MIT Center for Real Estate and MIT Media Lab City Science

Suffolk, the MIT Center for Real Estate and the MIT Media Lab City Science group published “Construction in the Age of AI: An Industry White Paper and Research Roadmap” on September 16, 2026. Its headline figure comes from a model Suffolk built on one multifamily project in San Francisco. By that model, AI applied across six parts of project delivery would have cut total cost by 17% to 20% and total schedule by 22% to 25%.

The workflows in scope run from permit applications and drawing review through schedules, RFIs, submittals and procurement. The authors describe their evidence as “suggestive rather than definitive,” and the methodology appendix says it cannot yet support causal conclusions.

What it does

Reported maximums reach 39% on time and 21% on cost, mostly from early pilots

The paper sorts AI uses into six domains, which the authors call levers: design automation, offsite manufacturing, permitting, scheduling, skilled labor and subcontracting, and supply chain and procurement. For each one it collects reported savings from a literature review of 178 sources, plus case studies, expert interviews and survey responses. A roundtable held on March 11, 2026, with more than 50 industry participants, settled the list. The acknowledgments name participants from Procore, Autodesk, Oracle and Bluebeam.

The percentages below are the maximums the authors found for each domain. Most come from early-stage pilots, and the sources include industries next to construction. A footnote in the paper says to read them as directional, and warns that they are not benchmarks validated on construction projects. The design automation figures also apply to a narrower base than the others, as the gray line under each name shows.

Maximum reported gain by AI domain, time in blue and cost in red

0%10%20%30%40%Design automationDesign cycle time; design and engineering cost39% time21% costOffsite manufacturingProject timeline; total cost32%14%PermittingProject time; cost24%8%SchedulingProject timeline; total cost18%11%Skilled labor and subcontractingProject time; total cost17%13%Supply chain and procurementTotal project time; total cost15%10% 0%20%40%Design automationDesign cycle time; design and engineering cost39% time21% costOffsite manufacturingProject timeline; total cost32%14%PermittingProject time; cost24%8%SchedulingProject timeline; total cost18%11%Skilled labor and subcontractingProject time; total cost17%13%Supply chain and procurementTotal project time; total cost15%10%
Maximum reported time reduction and cost savings, percent, from the white paper’s literature review, with what each percentage applies to. Directional figures, not validated on construction projects. Source: Suffolk and MIT white paper. Download the data (CSV)

The readers who get the most from the paper are preconstruction leads, VDC and scheduling managers, and CFOs who have to justify an AI budget. The domain sections name tools already sold to contractors. They cite ALICE Technologies and nPlan for schedule optimization, OpenSpace for tracking installed work against the plan, and TrunkTools’ TrunkRFI agent for flagging unnecessary RFIs and drafting needed ones. Procore’s in-platform agents come up for RFI deflection and drafting, and LightTable and BuildCheck for drawing review. (The paper identifies nPlan as a Suffolk Technologies portfolio company.)

The sample project

Design has the model’s largest schedule cut, 7 of 24 months

Suffolk then took one completed job and estimated what the six levers would have done to it together. Section 5 of the paper describes a 180,000-square-foot multifamily development finished in 2024. Table 1 breaks the result down by phase.

Modeled AI savings on the sample project, by phase.
PhaseBaseline
cost ($M)
Modeled
cut ($M)
Cost barCost
cut
Baseline
months
Modeled
cut (mo.)
Time barTime
cut
Acquisition2700%300%
Entitlement1.40.054%11219%
Design61.220%24730%
Preconstruction30.310%9114%
Construction1282822%30714%
Finance71.217%no figureno figureno figure
Cost escalation71.217%no figureno figureno figure
Total1793218%511122%

Red marks the modeled cost cut and blue the modeled schedule cut, each drawn at the end of the phase’s baseline bar. Cost bars share one scale, and time bars share another. Dollars in millions, time in months. Phase months overlap, so they do not add to the 51-month total. Source: Suffolk and MIT white paper, Table 1. Download the data (CSV)

Start at the bottom row. The model takes $32 million off a $179 million baseline, which is 18%, and 11 months off a 51-month timeline, which is 22%. Most of the dollars come from construction, where $28 million of the $32 million sits against a $128 million phase. Design has the largest schedule cut, 7 of 24 months. The model leaves acquisition untouched. The months do not add to the total because phases overlap, and the paper notes this under its table.

The authors then run the savings through to developer returns. By Suffolk’s internal model:

+5 to 6 points

Unlevered IRR, the project’s return before any debt, for example from 15-20% to 20-25%

+1 to 2 points

Yield on cost

Evidence so far

No project has recorded the 17% to 20% savings

The authors do not claim an independent measurement of AI on any construction project, and no job has recorded the 17% to 20% savings. The appendix calls the paper a methodology statement, and it lists what a real test would need: project-level data covering several hundred projects, and comparisons between AI-assisted and conventionally managed jobs that control for complexity, geography, delivery method and owner type.

The per-domain figures come from other parties, and the paper attributes them. Volumetric Building Companies reports up to 47% schedule savings from offsite manufacturing. The paper cites a Deloitte estimate that AI-assisted estimating cuts budget and timeline deviations by 10% to 20% and engineering hours by 10% to 30%. In its passage on ALICE Technologies, the paper says AI planning tools cut costs by up to 30% on one large solar farm, and it does not name the project or its baseline.

The paper gives the sample project’s baseline as $150 million, $179 million and $180 million

Suffolk calls its model a first pass, and some figures in the paper disagree with each other. If you plan to quote the paper in a budget request, check these first:

BASELINE COSTSample project, $ million140150160170180190$150MFigure 5$179MTable 1$180Msummary footnote DURATIONInitial design to closeout, years44.254.54.754.25 yearsTable 1, 51 months4.5 yearsSection 5 SCHEDULINGPercent, as stated in two placesCOST SAVINGS8%section11%summary tableSCHEDULE CUT15%model uses18%literature0%10%20%
  1. Baseline cost. The footnote under the summary table anchors survey responses to a $180 million project. Figure 5 labels the sample project $150 million. Table 1 totals $179 million.
  2. Duration. Section 5 gives 4.5 years from initial design to closeout. Table 1 gives 51 months, which is 4.25 years.
  3. Scheduling savings. The summary table gives up to 11% cost savings for scheduling, and the scheduling section gives 8%. A footnote there says the model uses a 15% schedule reduction, where the literature reports up to 18%.

Each red line spans two figures the paper gives for the same quantity. Source: Suffolk and MIT white paper

Expert ranking

Scheduling ranked first, with 163 points, once each group’s votes for its own lever came out

For a contractor choosing where to start, the expert ranking says more than the model. Participants scored each lever by Borda count, a ranking method that awards points by where an item falls on each voter’s list. Design automation collected the most points, 206, but 89 of them came from the design group voting for its own lever. With each group’s votes for its own lever removed, scheduling ranked first with 163 points.

The design group gave its own lever 89 points, the largest single vote in the matrix

DESIGNOFFSITESUPPLYLABORSCHEDULEPERMITDesign automation892027242620Offsite manufacturing512916191214Supply chain451816262816Skilled labor and subcontracting331420231620Schedule optimization691528232828Permitting41912201037POINTS FROM EACH EXPERT GROUPTOTAL POINTS05010015020025020689163 DesignOffsiteSupplyLaborSched.PermitDesign892027242620Offsite512916191214Supply451816262816Labor331420231620Schedule691528232828Permit41912201037POINTS FROM EACH GROUPTOTAL POINTSDesignOffsiteSupplyLaborSchedulePermit010020020689163
Borda points by AI lever (rows) from each expert group (columns). Darker cells hold more points, and the red rings mark each group’s votes for its own lever. In the totals, the light segment is those own-group votes. Source: Suffolk and MIT white paper. Download the data (CSV)

The skilled labor and subcontracting group spread its points almost evenly, between 19 and 26 per lever, which the authors read as a sign that field productivity depends on work done upstream.

How to measure it on your projects

Measure forecast finish slip at every monthly schedule update

So the schedule is where to start measuring. Track forecast finish slip, defined as the calendar days between the baseline substantial completion date and the forecast date in each monthly schedule update. If an AI scheduling or progress-tracking tool helps, slip on the pilot job should grow more slowly than on comparable jobs without it.

Capture the baseline before the pilot starts. Pull the monthly updates from three to five completed jobs of similar type and size, and record the slip at each update and at completion. The white paper asks owners to fund pilots with outcome measurement written into the project brief from day one, and a contractor’s own trial needs the same setup.

First step on one jobExport every monthly update from the P6 or Microsoft Project file for your most recently completed project, and chart the forecast completion date by month in Excel or Power BI.

Availability and cost

The white paper is free, as a PDF linked from Suffolk’s news page. The MIT research team is designing data collection instruments for the next phase. Firms that want to contribute project data, take part in surveys or join the research advisory panel can register through the MIT Center for Real Estate. The paper gives no date for that phase.