What a Preliminary Engineering Report Actually Is
A Preliminary Engineering Report (PER) is a pre-design study that establishes whether an infrastructure project is feasible, affordable, and sound— defining scope, an Engineer's Opinion of Probable Cost, schedule, and funding strategy before any design begins. It's the document that decides whether a project is worth designing at all.
The structure isn't improvised. A PER follows a standard, agency-defined outline, and the most widely used version comes from the USDA Rural Utilities Service Bulletin 1780-22, the de-facto national standard for municipal water and wastewater projects. Most reports track these sections:
- Executive Summary
- Project Planning Area
- Existing Facilities
- Need for Project
- Alternatives Considered
- Proposed Project (the selected alternative, with cost estimate, schedule, and permits)
- Conclusions & Recommendations
Here's the part firms forget under deadline pressure: a PER develops scope, cost, and schedule, but it is not the project's design3. It's the feasibility case, not the construction documents. Section names and counts shift by sector, but the USDA RUS structure is the anchor most agencies recognize.
Knowing the sections matters less than knowing which of them a licensed engineer's seal actually has to stand behind.
What a PE's Seal Certifies— and Why It Sets the Delegation Line
A professional engineer's seal certifies that the work was prepared by the engineer or under their direct supervision, and that the engineer is personally accountable for everything in the sealed document4. That single fact is what makes delegation possible. Junior staff can prepare sections; only a PE can seal them.
State licensing boards are explicit about this. In New York, reports prepared by the engineer or by a subordinate under their supervision must be signed and sealed when filed with public officials5. The subordinate's work is lawful. The seal is what makes it official.
And the whole report doesn't get stamped page by page. For large multi-section documents, Florida's board notes that typically only the index or cover sheet is sealed, with the PE indicating the sections they're in responsible charge of6. Responsible charge is the standard that governs all of this— it means the PE actually directed and reviewed the work, not that they glanced at a finished file and signed it.
So the accountability is total, even when the keystrokes aren't the PE's. That's the line that matters: a junior engineer or an AI tool can build a section, but the senior engineer's signature is a personal, legal warranty on the entire report. Treat that warranty as the starting point, not an obstacle— and start on the side of the line that can't move: the sections only a senior engineer can own.
The 3 PER Sections That Require a Senior Engineer
Three parts of a PER require a senior engineer's professional judgment— and the seal that stands behind them: the alternatives analysis and selection, the preliminary design criteria and sizing, and the conclusions and recommendations. No junior draft or AI tool takes these off a senior engineer's plate, because what they require isn't production. It's accountable, licensed judgment.
A note on where this split comes from: this is our framework, not a regulatory list. No agency publishes "the three sections a PE must write." What the law fixes is the seal and responsible charge4; the judgment-versus-production mapping is our reasoned read of how that accountability works in practice.
- Alternatives analysis & selection. Weighing trade-offs— cost, risk, operability, life-cycle impact— and choosing the recommended path is judgment. A junior can assemble the options matrix; deciding which alternative the firm will stand behind is the engineer's call.
- Preliminary design criteria & sizing. These are the assumptions that drive everything downstream: design flows, loading, capacity, code basis. Get them wrong and the whole report is wrong.
- Conclusions & recommendations. This is what the seal explicitly stands behind— the engineer telling a funding agency "this is sound," in their own professional name.
There's a fourth non-delegable element that isn't a section at all: the seal itself. Even the cost estimate has a judgment core— setting the unit-cost assumptions and contingencies is senior work, even when juniors build the tables. More on that next.
Everything else— the data, the compilation, the narrative— is where the firm is overspending senior hours.
The 7 PER Sections That Don't Need a Senior Engineer
Seven parts of a Preliminary Engineering Report are data-gathering, compilation, or narrative work that junior staff and AI can take to a near-final draft— leaving the PE to review, set assumptions, and seal rather than produce from scratch. Again, this is our framework for splitting the work, not a regulatory line. The section names come straight from the canonical structure27; the call on what's delegable is ours.
- Project Planning Area. Demographics, service-area maps, and growth projections. It's research and presentation— pull the data, map the boundaries, summarize the trends. No seal required to count households.
- Existing Facilities inventory & history. Record drawings, condition and inspection records, capacity data. A junior assembling the as-built history is doing librarian's work, not engineering judgment. (If your team rebuilds this from scratch every time, your PER template may be older than your junior engineers.)
- Permit & regulatory requirements list. Cataloging applicable permits, approvals, and agency touchpoints. This is structured research with a clear right answer, which is exactly what juniors and AI do well.
- Environmental & sustainability considerations narrative. Assembling the impact and considerations write-up from known sources. The findings come from data; the narrative is compilation.
- Cost-estimate tabulation (Engineer's Opinion of Probable Cost). Here's the caveat that keeps this honest. The PE sets the assumptions— unit costs, contingencies, escalation— but juniors and AI can build, format, and check the tables once those numbers are fixed. The tabulation is delegable. The assumptions are not.
- Executive Summary. Written last, assembled from the finished sections. Summarizing decisions that are already made is compilation work, and it's the kind of writing AI is genuinely good at.
- Front & back matter. Title page, table of contents, references, figure assembly. Cover sheets and front matter aren't engineering work, yet they routinely eat senior hours on formatting no license requires.
Delegating production only saves senior time if the review model is real— not a rubber stamp.
The Review Model That Keeps Delegation Compliant
Delegation moves production off the senior engineer. It does not move responsibility. The PE still seals the report, so the savings are real only when supervision is genuine: clear assumptions set up front, structured review at defined checkpoints, and never a signature on work the PE didn't actually direct.
Does delegation create review burden? It can. A sloppy junior or AI draft can increase a senior engineer's workload instead of reducing it. If a PE has to untangle a bad alternatives matrix or re-derive cost assumptions someone guessed at, delegation has cost them time, not saved it. That failure mode is real, and it's why over-delegating without supervision is worse than not delegating at all.
The fix is sequencing. The PE sets the criteria and assumptions before anyone drafts, so juniors and AI work inside guardrails instead of inventing them. Then review happens at defined checkpoints— after the alternatives matrix, after the cost basis, after the full draft— rather than as one surprise at the end. A structured pre-QC pass that actually saves senior-engineer time is the difference between leverage and liability.
Responsible-charge standards vary by state, and the PE must genuinely supervise: direct the work, review it, and stand behind it. Do that, and the seal means exactly what it always meant. Skip it, and you've built a faster way to produce a report nobody should sign.
With the review model in place, AI becomes a multiplier on exactly the delegable sections— and nothing more.
Where AI Fits in PER Production Today
AI fits the same sections junior staff can draft— narrative, compilation, and boilerplate— and stops exactly where engineering judgment begins. It can populate templates, search decades of prior reports, and draft narrative sections. It can't make the alternatives, design, and recommendation calls a PE's seal certifies10.
The adoption picture is uneven, and that's the opportunity. Bluebeam's 2026 AEC Technology Outlook found that only 27% of AEC firms use AI for automation, problem-solving, or decision-making— yet the firms that have moved are seeing real returns8. The numbers behind that gap:
- 27% of AEC firms use AI for automation, problem-solving, or decision-making (Bluebeam 2026)8
- 68% of early adopters have saved at least $50,000 (Bluebeam 2026)8
- 46% of early adopters have saved 500–1,000 hours (Bluebeam 2026)8
- 63% of firms have an AI strategy in place or in development (ACEC 2025)9
- 41% of firms report labor shortages cause project delays (Newforma 2025)1
Read together, the case makes itself: the labor constraint is real, strategy interest is climbing, and early movers are saving hundreds of hours89. The work AI takes on is narrative, compilation, and boilerplate— the seven delegable sections of a Preliminary Engineering Report, not the three that need a seal.
One distinction matters. General-purpose chatbots can hallucinate technical specifics, and a fabricated ASTM reference or an invented capacity figure has no place in a sealed document. Report-specific (vertical) AI tools, trained on a firm's own prior reports and templates, fit PER work better than a general model writing from a blank page10.
The pattern shows up outside engineering, too. Fielding Jezreel, a federal grant-writing consultant, built an AI tool to draft his budget narratives— the formulaic writing he'd rather not do by hand— while keeping the judgment calls for himself. As he puts it, the tool "doesn't replace a grant writer"; the value is in the pairing, where a domain expert directs AI through work neither does as well alone. A PER's narrative sections are the same kind of formulaic work.
A few questions come up every time a firm tries to draw this line.
Frequently Asked Questions (FAQ)
What are the sections of a PER?
A PER typically contains an Executive Summary, Project Planning Area, Existing Facilities, Need for Project, Alternatives Considered, Proposed Project (the selected alternative with cost estimate, schedule, and permits), and Conclusions & Recommendations2. Names and counts vary by agency and sector, but the USDA RUS Bulletin 1780-2 format is the de-facto national standard.
Does a junior engineer need a PE to write a PER?
Junior staff and EITs (engineers-in-training) can prepare PER sections, but a licensed PE must review, seal, and take responsibility for the report5. The seal certifies the work was done by the engineer or under their direct supervision4. The preparation can be delegated; the accountability cannot.
Does every page of a PER need a PE stamp?
No. For large multi-section reports, typically only the index or cover sheet is sealed, with the PE indicating the sections they're in responsible charge of6. The seal covers the whole report through that index, not through a stamp on every page.
Can AI write a PER?
AI can draft the narrative, compilation, and boilerplate sections of a PER, but it can't make the engineering judgments— alternatives selection, design criteria, and recommendations— that the PE's seal certifies4. Vertical AI tools trained on prior reports fit this work better than general chatbots, which can hallucinate technical specifics10.
Free Your Senior Engineers for the Work Only They Can Do
The win is simple: senior engineers spend their hours only where their license actually matters. "AI writes your reports" was never the point.
Reclaim that capacity by moving the seven delegable PER sections— planning area, existing facilities, permits, environmental narrative, cost tabulation, executive summary, and front matter— to junior staff and AI, and reserving the seal for the three that earn it. The blocker is rarely capability. It's the templates, SOPs, and review workflow that make delegation safe and repeatable instead of risky.
The real work is building the review model and the report-production system around it. Buying an AI tool is the easy part. Firms that want help building the report-production workflows that make delegation safe usually don't need more engineers— they need their senior engineers pointed at the judgment only they can sign for. Put the seal where it belongs, and the Preliminary Engineering Report stops being a senior-engineer tax.
References
- Newforma, "Labor Shortages in AEC: The Hidden Impact and How Smart Firms Are Adapting (2025 AECO Project and Information Management Survey)" (2025) — https://www.newforma.com/labor-shortages-in-aec-the-hidden-impact-and-how-smart-firms-are-adapting/
- USDA Rural Utilities Service, "Bulletin 1780-2: Preliminary Engineering Report Guide" (current) — https://www.rd.usda.gov/files/UWP_Bulletin_1780-2.pdf
- Wessler Engineering / Morrison-Maierle, "Developing a Preliminary Engineering Report (PER)" (current) — https://info.wesslerengineering.com/blog/developing-a-preliminary-engineering-report-per
- National Society of Professional Engineers, "What a PE Says with their Signature and Stamp" (current) — https://www.nspe.org/sites/default/files/resources/pdfs/GR/NSPE_COPA_Stamp_Doc.pdf
- New York State Education Department, Office of the Professions, "Professional Seals and Signatures (Engineering Practice Guideline 3)" (current) — https://www.op.nysed.gov/professions/engineering/professional-practice/professional-seals-and-signatures
- Florida Board of Professional Engineers, "Signing and Sealing Engineering Documents" (2020) — https://fbpe.org/legal/signing-and-sealing-engineering-documents/
- Commonwealth of Virginia, "9VAC25-790-940. Preliminary Engineering Report" (current) — https://law.lis.virginia.gov/admincode/title9/agency25/chapter790/section940/
- Bluebeam, "Building the Future: Bluebeam AEC Technology Outlook 2026" (2025) — https://press.bluebeam.com/2025/10/new-bluebeam-report-shows-early-ai-adopters-in-aec-seeing-significant-roi-despite-uneven-adoption/
- American Council of Engineering Companies, "A Primer on AI Integration for Engineering Firms" (2025) — https://www.acec.org/wp-content/uploads/2025/10/AI-Adoption-in-AEC-Summary_Primer_October_2025.pdf
- AECbytes, "Knowledge Amplified: How AI is Revolutionizing Technical Report Deliverables in Engineering (Viewpoint #124)" (2025) — https://www.aecbytes.com/viewpoint/2025/issue_124.html