Turn your electrical drawings into reviewed, bid-ready device counts and measured runs—without giving up estimator control. Bring your own plan set to a guided demo and see exactly how the workflow fits your symbols, scope and estimating process.
Electrical takeoff software helps contractors turn electrical drawings into accurate, reviewable quantity takeoffs faster. It identifies and organizes device counts, lighting fixtures, panels, equipment and measured runs—reducing repetitive sheet-by-sheet markup while giving estimators clear quantities to validate.
Instead of replacing electrical estimators, AI-assisted takeoff software speeds up the most time-consuming measurement work. Estimators remain in control of symbol mapping, plan scale, scope decisions, revisions and final approval, helping teams prepare bids with greater consistency and fewer missed items.
Upload electrical plans, review detected quantities and move bid-ready takeoff data into your estimating workflow. See the process on your own drawings and evaluate how it can shorten takeoff time without sacrificing accuracy or accountability.
Receptacles, switches, lighting fixtures, junction boxes and similar repeated symbols — the items that dominate an electrical takeoff by sheer count and consume the most manual marking time.
Conduit, cable tray, homeruns and similar linear items measured along the drawing against a verified plan scale. A plan-view distance is not the same as installed cable length, which matters when the number reaches pricing.
Panelboards, switchgear, transformers and equipment that usually appear once per location and are reconciled against the schedules rather than counted off the plan view alone.
Data, fire alarm, security and controls devices, which often carry their own legend and are drawn on separate sheets — a common place for a takeoff to miss scope entirely.
The device legend and fixture schedule are where a drawing says what its symbols mean. Any automated detection is only as good as its mapping onto the legend actually used on your sheets.
The output is a structured quantity list for your estimating process. Material pricing, labor units, assemblies and markup are not produced here — see the comparison table below.
Five stages from an electrical plan set to quantities an estimator has signed off. Which file types, device classes and outputs apply to your projects is established against your own drawings in the demo — one click does not produce a priced bid, and no stage here claims to.
Bring the electrical sheets for a job you have already taken off by hand, so there is a known answer to compare against. Whether a sheet is a clean vector export or a scan of a scan changes what is possible, and that is the first thing we look at.
The device legend on your drawing defines what each symbol means. Detection is mapped onto that legend rather than a generic symbol set, and anything the model is unsure about is surfaced instead of being quietly folded into a total.
Located items become structured quantities grouped into electrical categories — devices counted, runs measured, equipment reconciled against schedules — each retaining the sheet and measurement it came from.
Every quantity is presented for an estimator to accept, adjust or reject, with low-confidence items flagged first. Addenda and reissued sheets are a routine part of electrical bidding, so revision handling is one of the things worth testing in the demo.
Approved quantities are prepared as a structured list for the next step in your estimating process. Which export routes and system interfaces apply to you is confirmed during evaluation rather than listed here as shipping connectors.
The six things that decide whether an electrical takeoff tool survives a bid week. We demonstrate each on your own sheets rather than listing them as specifications — a feature that works on a clean sample drawing and one that works on a reissued low-voltage sheet are not the same feature.
Your device legend is the reference the whole takeoff depends on. Detection is mapped onto the symbols your design office actually draws, and custom or non-standard callouts are exactly the case to put in front of a demo — they are where generic tools lose counts quietly.
Receptacles, switches, lighting fixtures, junction boxes and similar repeated items counted per class rather than as one total, so a symbol that appears rarely cannot hide inside a headline number.
Every measured length depends on scale being right for that sheet. A wrong scale does not produce an obvious error — it produces confident, plausible, wrong footage. Scale is established and verified per sheet before measurement.
Conduit, cable tray and homerun runs measured along the drawing. Worth understanding clearly: a plan-view distance is a 2D measurement, not an installed cable length with drops, slack and routing allowances.
Accept, adjust or reject each item, with notes against the drawing and confidence shown rather than hidden. How quickly an estimator can confirm or correct output is the measure that decides adoption.
Sheets organized by project with reissued drawings kept distinct. Addenda are routine in electrical bidding, and a takeoff that cannot tell one revision from another creates rework instead of removing it.
Electrical symbols, schedules and plan notes vary between design offices and often between sheets on the same job, so support is confirmed per project rather than declared in advance. Every row below reads Confirm during demo for exactly that reason — we would rather show you than promise you.
| Electrical element | Typical measurement | Product support status | Estimator verification |
|---|---|---|---|
| Receptacles, switches, devices | Count by device class | Confirm during demo | Spot-check counts per class against a sheet taken off by hand |
| Lighting fixtures | Count by fixture type | Confirm during demo | Reconcile against the fixture schedule, not the plan view alone |
| Conduit and cable tray | Linear measurement | Confirm during demo | Check scale per sheet; 2D length is not installed length |
| Homeruns and circuit routing | Linear measurement / count | Confirm during demo | Confirm routing assumptions before the figure reaches pricing |
| Panelboards and equipment | Count / schedule reconciliation | Confirm during demo | Verify against panel schedules and single-line drawings |
| Low-voltage, data, fire alarm | Count by device class | Confirm during demo | Confirm the separate legend used on those sheets is read correctly |
| Boxes, fittings, supports | Count / derived quantity | Confirm during demo | Establish whether these are drawn, scheduled or assumed by rule |
| Material pricing and labor units | Not a takeoff output | Not in scope | Belongs to your estimating process — see the comparison below |
A takeoff is a measurement input. An estimate is a commercial output. Tools that do both exist; this one is a takeoff tool, and the column on the right is the honest scope line.
| Task | Electrical takeoff | Full electrical estimating | DreamzTech verified scope |
|---|---|---|---|
| Plan-based quantities | Core function | Included as an input | Demonstrated on your drawings |
| Symbol and device counts | Core function | Usually included | Demonstrated on your drawings |
| Length measurement | Core function | Usually included | Demonstrated on your drawings |
| Materials pricing | Not included | Core function | Not in scope |
| Labor units | Not included | Core function | Not in scope |
| Assemblies | Not included | Core function | Not in scope — discuss as custom development |
| Overhead and markup | Not included | Core function | Not in scope |
| Bid proposal output | Not included | Core function | Not in scope |
| Revisions and approval | Drawing revision handling | Bid revision control | Confirm during demo |
If you need pricing, labor units and assemblies joined to the takeoff, that is an engineering scope rather than a product setting — see custom AI takeoff development.
Electrical drawings carry ambiguous symbols, mixed revision sets, project-specific legends and plenty of low-resolution scans. Any AI pass over them produces a reviewable first draft, not a finished takeoff — and a tool that pretends otherwise just moves the error somewhere harder to find.
Four things about electrical plan sets that decide how a takeoff goes. Each names what can reasonably be automated and what stays a manual project control — the distinction matters more on electrical work than on most trades.
The legend sheet defines what every symbol means on that job. Where a design office uses custom device callouts, a generic symbol library will not recognize them, and that gap is the most common reason an automated electrical count comes back wrong.
Fixture schedules, panel schedules and general notes carry quantities and exclusions that never appear as symbols. Reflected ceiling plans and riser diagrams hold scope the power plan does not show.
A reissued sheet can invalidate work already done, and finding what actually changed between versions is slow and error-prone by eye. Revision handling is worth testing with your own reissued drawings.
A clean vector export and a photographed scan are different problems. Resolution, skew and linework clarity cap what detection can achieve regardless of the software, and scale must be verified per sheet.
Electrical takeoff rewards automation because the same symbols repeat across every sheet of every job. These are the four roles that get the most out of an evaluation.
Problem: multi-sheet device counts on commercial jobs eat the bid window, and capacity is capped by how many estimators you can put on a plan set. Outcome: the repetitive marking shrinks, so the limit becomes review time rather than measuring time.
Problem: high bid volume on smaller jobs, where the time per takeoff matters more than the complexity of any one. Outcome: a faster repeatable pass over standard plan types, with the estimator confirming rather than marking from scratch.
Problem: electrical is one of several disciplines to quantify, each with its own legend and conventions. Outcome: consistent method and one audit trail across the trade instead of per-estimator spreadsheets nobody else can reconstruct. Bidding plumbing too? See plumbing takeoff software.
Problem: validating a subcontractor's electrical quantities without re-doing the entire takeoff by hand. Outcome: an independent count to compare against, with the derivation visible behind each number.
What we can actually evidence: custom application engineering, computer vision work, integration delivery into enterprise systems, and a demo that runs on your own electrical sheets before anyone asks you to commit. If the product route does not fit your legends and revision process, we can build the workflow instead — and we will say which. No accuracy percentage, speed multiple or competitor comparison appears on this page, because none has a published methodology behind it.
Review the method, worked example and standards behind a reliable electrical quantity takeoff. See how estimator checks turn device counts and measured runs into evidence your team can trust before bid submission.
Three representative electrical projects show the quantities reviewed, the drawing risks checked and the evidence retained before bid preparation.
Takeoff scope: Lighting, receptacles, switches, junction boxes and measured conduit or cable-tray runs.
Estimator evidence: Estimator review reconciles legend classes, flags missed symbols and records every count or length correction.
Takeoff scope: Unit devices, common-area lighting, panels and feeders across repeated floor plans.
Estimator evidence: Typical-unit logic is checked floor by floor, while panel and equipment schedules prevent omitted scope.
Takeoff scope: High-bay fixtures, devices, tray or conduit routes, fire-alarm and data symbols.
Estimator evidence: System-specific legends and revision checks separate similar symbols and prevent duplicated quantities.
DreamzTech delivers AI-assisted electrical takeoff capabilities for contractors who need faster device counts and measured runs without losing estimator control. The workflow connects legends, plans, schedules, review decisions and structured output in one traceable process.
Project-specific device legends are mapped before detection so receptacles, switches, fixtures and low-voltage symbols are classified in context.
The workflow locates electrical devices and lighting classes across multi-sheet plan sets while retaining every detected position for review.
Conduit, cable tray and other plan runs can be measured after per-sheet scale calibration, with assumptions visible to the estimator.
Panelboards, equipment and fixture schedules are checked against plan detections to surface missing or conflicting scope.
Low-confidence items, false positives, missed symbols and quantity corrections are reviewed before any result moves into estimating.
Approved counts and lengths are organized by class, sheet and revision for spreadsheet, assembly mapping or scoped system integration.
We run the demo on your drawing, not a recorded one. A prepared sample sheet with a tidy legend proves nothing about your reissued low-voltage plans — so bring a job you have already taken off by hand and we will work through it live with your estimators in the room.
Share a representative electrical plan set ahead of the session, ideally one already taken off manually so there is a known answer to compare against. Include the device legend and, if you have one, a reissued sheet — revisions are where takeoff tools usually struggle. We check format, resolution and scale handling beforehand so the call is spent on substance.
We run the workflow on your sheets and work through the output as an estimator would: which devices were found, which were missed, where confidence is low, how the legend was interpreted, and how long it takes a person to confirm or correct each item. Your estimators should be the ones judging this part.
You get an honest read: where it performed, where it did not, what would need building for your legends or revision process, and whether a product evaluation or a custom build makes more sense. If neither is worth it for the drawings you bid, we will tell you that.
Two routes, genuinely different commitments. The demo is how you find out which one your drawings and estimating process point to — and either answer is a fine result.
| Need | Product evaluation this page | Custom AI development |
|---|---|---|
| Common takeoff workflow | Evaluate the product against your plan sets | Unnecessary — if the product fits, use it |
| Proprietary symbol sets | Confirm coverage of your legend in the demo | Trained on your own device callouts and legend conventions |
| Enterprise integrations | Export and interfaces confirmed during evaluation | Engineered to your target systems, bounded by their APIs |
| Fully bespoke estimation process | Out of scope — this is a takeoff tool | Pricing, labor units and assemblies built into the workflow |
| Source-code / IP ownership | Licensed use under the product agreement | Defined in the project agreement, subject to third-party component terms |
| Time to first value | Shorter — evaluate, then adopt if it fits | Longer — discovery and build before first use |
Licensing a product is not the same as owning it. Source-code and model ownership terms belong to a development agreement, not a software licence — if owning the capability matters, that points to custom AI takeoff development. For the cross-trade workflow, see AI takeoff software for construction.
Every electrical takeoff we produce runs through the same six-stage protocol, and each stage names what the software does, what a human verifies, and what gets written down. The protocol is published here so a reviewer can check our work against it rather than take an accuracy claim on trust.
Electrical takeoff goes wrong in predictable places: a legend that reuses a symbol, a revision cloud that changes a device type without changing its symbol, and sheets that overlap at a match line. The protocol below exists to put a named check in front of each of those failures.
Every sheet in the set is logged with its number, revision, date and scale before anything is measured. Sheets that carry a revision later than the set’s issue date are flagged, not silently used.
Review gateAn estimator confirms the sheet register matches the drawing index. A set that is missing a sheet listed in its own index does not proceed to measurement.
The symbol legend is read first and turned into an explicit symbol-to-item map for that project. Where a project reuses a generic symbol for two different devices, the map records the distinction and the rule for telling them apart — usually a tag, a callout or a circuit number.
Review gateNo symbol class is counted until it has a legend entry. Unmapped symbols are raised as a question to the design office rather than guessed.
The model scans each sheet for the mapped symbol classes and returns counts with a location for every instance, so each number can be traced back to a point on a drawing.
Detection produces candidates, not quantities. Nothing from this stage reaches an export untouched.
Raceway, feeder and branch runs are measured against the sheet’s stated scale, with the scale verified against a known dimension on the sheet rather than trusted from the title block.
Review gateWhere a measured scale disagrees with the title block, the sheet is set aside and the discrepancy reported. Drawings are frequently reduced in printing or PDF export, and a 4% scale error is a 4% error on every linear quantity on that sheet.
A human estimator reviews each symbol class against the drawing, working from the lowest-confidence classes upward. Corrections are recorded as corrections — the original detected figure, the corrected figure, the sheet, and who changed it.
Review gateThis is the stage that decides the number. Review is not a spot check; the class-level audit trail is what the exported quantity set is built from.
Quantities are grouped under CSI MasterFormat Division 26 headings and exported with the sheet reference, the revision they were taken from, and the correction log attached.
Review gateThe export states which drawing revision it reflects. A quantity set with no revision stamp is not a deliverable, because the next revision will change it.
This is a complete electrical takeoff worked end to end on a synthetic drawing set — a two-storey office fit-out of 18,400 sq ft, prepared for this page so the arithmetic can be published in full. It is not a client project and the numbers are not a performance claim.
The set comprises 14 sheets: 6 lighting plans, 5 power plans, 2 single-line diagrams and 1 panel schedule, all at 1/8 in. = 1 ft 0 in. except the enlarged electrical room at 1/4 in. scale.
| Device class (Division 26) | Count |
|---|---|
| Duplex receptacle, standard | 146 |
| Duplex receptacle, GFCI (wet areas) | 18 |
| Quad receptacle | 12 |
| Floor box receptacle | 5 |
| Single-pole switch | 54 |
| Three-way switch | 16 |
| Occupancy sensor | 22 |
| Luminaire, 2×4 LED troffer | 184 |
| Luminaire, 2×2 LED troffer | 46 |
| Downlight, 6 in. LED | 58 |
| Exit sign | 14 |
| Emergency light | 19 |
| Data outlet | 96 |
| Panelboard | 8 |
| Total counted devices | 698 |
| EMT conduit, trade size | Measured length (LF) |
|---|---|
| 1/2 in. | 4,820 |
| 3/4 in. | 2,140 |
| 1 in. | 680 |
| 1-1/4 in. | 240 |
| Total EMT | 7,880 LF |
Conductor is not counted on the drawing; it is derived from measured raceway and the circuit arrangement, then carried with its waste allowance stated rather than folded in silently.
| Conductor | Raceway (LF) | × conductors | Subtotal (LF) | +10% waste |
|---|---|---|---|---|
| #12 THHN in 1/2 in. EMT | 4,820 | 3 | 14,460 | 15,906 |
| #10 THHN in 3/4 in. EMT | 2,140 | 3 | 6,420 | 7,062 |
That correction is the whole argument for the review stage. The automated pass was not wrong about where the devices were — it found all 151 — it was wrong about what they were. A takeoff that reported a single accuracy percentage would have scored this sheet as correct.
Electrical takeoff is governed by published standards, not by house rules. These are the documents that determine what has to be counted, how items are classified on handoff, and how firm an estimate built on a given level of detail may be called.
| Standard or reference | Why it governs the takeoff |
|---|---|
| NFPA 70, National Electrical Code | Defines the device, circuit, conductor and raceway requirements that determine which electrical items a drawing set must show — and therefore what a takeoff has to count. |
| NEMA standards | Device, enclosure and configuration designations. Two symbols that look alike on a plan can carry different NEMA types and different unit costs, so the legend governs the count. |
| CSI MasterFormat | The classification system quantities are grouped under when they are handed to an estimator. Division 26 covers electrical, 22 plumbing, 23 HVAC. |
| CSI MasterFormat Numbers & Titles | The published division and section numbering used to label exported quantity groups. |
| AACE International Recommended Practices | Cost estimate classification. Defines how firm an estimate derived from a given level of takeoff detail may legitimately be called. |
| ASTM E2516, Standard Classification for Cost Estimate Classification System | A published classification for estimate maturity, used to state the confidence class of a quantity set rather than implying a single accuracy figure. |
Two of these matter more than the rest on a live bid. MasterFormat Division 26 decides how quantities are grouped when they leave us, which is what makes a takeoff usable by someone else’s estimating system. AACE’s estimate classification is what stops a quantity set from being described as more certain than its inputs allow.
Related trade takeoffs follow the same protocol with different symbol classes and codes: plumbing takeoff software and HVAC and mechanical takeoffs. The cross-trade workflow is described on the AI takeoff software page.
Page provenance. Method and worked example on this page are maintained by the DreamzTech construction technology team and were last reviewed on . The worked example uses a synthetic drawing set prepared for publication, not client work. Standards above are linked to their publishers; DreamzTech is not affiliated with and does not speak for any of them. To request the full written method, ask during a demo.
What electrical estimators ask when evaluating takeoff software: what it counts, how conduit lengths differ from installed cable, whether it produces a bid, how custom legends are handled, how accuracy is checked, and what to put in front of a demo.
Software that extracts quantities from electrical drawings — counts of devices and fixtures, measured lengths of runs, and the material items a job needs.
It is not the same as electrical estimating software, which adds material pricing, labor units, overhead and markup to turn those quantities into a bid. Takeoff is the measurement input; estimating is the commercial output. Treat a takeoff tool as something that removes repetitive marking from an estimator’s day, not something that produces a priced proposal.
Computer vision locates repeated symbols across the sheets, then an estimator reviews what it found.
As a class of capability, detection models are trained to recognize shapes on a drawing and return each one with a location and a confidence value. On electrical work the legend matters enormously, because device callouts differ between design offices. What this particular workflow detects on your sheets is established in a demo rather than claimed here — and anything it is unsure about should be surfaced for checking rather than folded silently into a total.
Those are exactly the items to test in a demo, and we do not claim coverage in advance.
Receptacles, switches and lighting fixtures are the highest-volume symbols on most electrical plans, which is why they are where automation pays off and also where a generic symbol library most often fails. Coverage depends on how your drawings represent them and whether your legend uses custom callouts. Bring a sheet you have already counted by hand and we will compare per device class.
It can measure distances on the drawing. That is not the same as installed cable length.
A 2D plan-view measurement follows the route as drawn. Installed quantities add vertical drops, slack at terminations, routing deviations around obstructions and whatever allowance your estimating standards apply. Any tool that reports a plan length as a cable quantity is skipping a judgment that belongs to your process. Expect measured lengths as an input and apply your own installation assumptions on top.
No.
A takeoff produces quantities. A bid needs material pricing, labor units, assemblies, overhead, margin and risk applied to those quantities, plus the commercial judgment that decides what to include and exclude. Some products combine takeoff and estimating; this is a takeoff workflow, and its output is designed to hand off into whichever estimating system you already price in. Joining the two is an engineering scope — see custom AI takeoff development.
This is the single most important thing to test, and the honest answer depends on your legend.
Electrical symbol conventions vary by design office and sometimes by project within the same office. A detector tuned to one firm’s callouts will underperform on another’s. Where a product cannot be configured far enough for your symbol set, training on your own annotated drawings is a custom development scope rather than a product setting. Put your most unusual legend in front of the demo — not your cleanest one.
Against a plan set you have already taken off manually, compared per item class.
We do not publish an accuracy percentage, because a figure without the drawings and the test method behind it is marketing rather than evidence. The meaningful check is: counts compared per device class so a rarely-used symbol cannot hide in an aggregate; measured lengths compared against known quantities with scale verified per sheet; and missed items looked for deliberately, since an undetected item appears nowhere to be questioned.
PDF is the normal input for this category; the useful question is what it does with your PDFs.
A vector export carries structure that makes detection considerably more reliable. A scanned or photographed sheet does not, and scan quality caps achievable results regardless of software. Revision handling matters just as much on electrical work, because addenda routinely invalidate completed takeoff. Supported formats and revision behaviour are confirmed against your own sheets during evaluation rather than listed here.
Structured export is the normal handoff route; specific connectors are confirmed before they are committed to scope.
We deliberately do not publish a list of prebuilt integrations with named estimating or ERP platforms. What a given system can accept depends on the interface it genuinely exposes, which is assessed during evaluation. Where a route needs building rather than configuring, that is custom development and is quoted separately. The part that usually takes longest is mapping detected items onto your own cost codes and assemblies.
Put your hardest drawings in front of it, not your cleanest.
Worth checking: whether your device legend and any custom callouts are read correctly; counts per device class against a sheet you already took off by hand; how conduit and homerun lengths are measured and what assumptions apply; whether low-voltage sheets with their own legend are handled; how a reissued drawing is compared against the previous revision; how long correction actually takes an estimator; and what the output looks like arriving in your estimating process.
Show us the electrical drawing types and quantity outputs you work with. We'll help determine whether the current software or a tailored implementation fits your workflow.
Send your electrical sheets, legend and the quantities you need. We will run the workflow on them in the session and give you a straight answer on fit — product, custom build, or neither.
Bring a plan set you have already taken off by hand, with its legend and a reissued sheet. We will show you where AI-assisted electrical takeoff helps and where it does not. No prepared sample sheets.
