SynthPID: P&ID digitization from Topology-Preserving Synthetic Data

Suraj Prasad; Pinak Mahapatra

cs.CVApr 15, 2026

SynthPID: P&ID digitization from Topology-Preserving Synthetic Data

Authors: Suraj Prasad, Pinak Mahapatra

Organizations: Indian Institute Of Technology Bombay

Abstract

Automating the digitization of Piping and Instrumentation Diagrams (P&IDs) into structured process graphs would unlock significant value in plant operations, yet progress is bottlenecked by a fundamental data problem: engineering drawings are proprietary, and the entire community shares a single public benchmark of just 12 annotated images. Prior attempts at synthetic augmentation have fallen short because template-based generators scatter symbols at random, producing graphs that bear little resemblance to real process plants and, accordingly, yield only approximately 33% edge detection accuracy under synth-only training. We argue the failure is structural rather than visual and address it by introducing SynthPID, a corpus of 665 synthetic P&IDs whose pipe topology is seeded directly from real drawings. Paired with a patch-based Relationformer adapted for high-resolution diagrams, a model trained on SynthPID alone achieves 63.8 +/- 3.1% edge mAP on PID2Graph OPEN100 without seeing a single real P&ID during training, closing within 8 pp of the real-data oracle. These gains hold up under a controlled comparison against the template-based regime, confirming that generation quality drives performance rather than model choice. A scaling study reveals that gains flatten beyond roughly 400 synthetic images, pointing to seed diversity as the binding constraint.

Figures & tables

Figure 1: Method overview. Twelve real P&IDs serve as structural seeds. Our generator perturbs each seed’s topology and re-renders it, producing SynthPID: 665 diverse synthetic P&IDs with full graph annotation. After connector collapsing and patch extraction, a Relationformer is trained on these patches and evaluated on the original real images via patch merging.

Figure 2: The visual domain gap. Real OPEN100 P&IDs (top) have white backgrounds and professional drafting conventions; SynthPID images (bottom) have a grey background and a different rendering style. Despite this, both share the same symbol vocabulary and, critically, the same pipe topology statistics.

Figure 3: Degree distribution (a) and edge density (b) for three corpora. SynthPID tracks the real OPEN100 distribution closely; template-based synthetic data does not.

Node classes (7)	Edge classes (2)
valve, pump, instrumentation	solid
general, tank, arrow, inlet/outlet	non-solid

Table 1: Output vocabulary used throughout all experiments.

Dataset	# Images	Nodes/img	Edges/img
OPEN100 (real)	12	$147\!\pm\!52$	$65\!\pm\!28$
SynthPID (seed-synth)	665	$231\!\pm\!44$	$152\!\pm\!37$

Table 2: Dataset statistics after connector collapsing.

Method	Train Data	Node [email protected]	Edge mAP
Reference [ 18 ]
Modular Digitization	60 real P&IDs	52.14	45.89
Synth-only †	2,000 template	$\sim$ 41	$\sim$ 33
Ours
E1: Oracle	12 real, CV	67.8 $\!\pm\!$ 3.2	69.4 $\!\pm\!$ 3.8
E2: SynthPID	665 synth, 0 real	64.2 $\!\pm\!$ 2.6	63.8 $\!\pm\!$ 3.1

Table 3: Node [email protected] and edge mAP on OPEN100 (stitched evaluation). Bold: our primary result (E2, zero real training data).

\dagger

Estimated from Appendix A5 of [ 18 ] , described as “suffers significantly” relative to the full model. Our results: mean

\pm

std over 15 runs.

Panel of Figure 4 — Figure 4: Edge mAP versus number of synthetic training images. Performance rises sharply up to around 400 images (+12.5 pp from 100) and then begins to level off (+5.1 pp from 400 to 665), suggesting that seed diversity rather than image volume is the limiting factor.

Class	[email protected]	Class	[email protected]
valve	74.3	tank	77.6
pump	68.2	arrow	52.4
instrumentation	61.7	inlet/outlet	65.8
general	48.3	mean	64.2

Table 4: Per-class node [email protected] under E2 (zero real training data).

Figure 5: Ground truth (left) versus SynthPID predictions (right) on a real OPEN100 P&ID (top row) and a synthetic image (bottom row), both under E2. Solid boxes and solid edge lines are true positives; dashed elements are false positives. The model reconstructs the main process topology without any real training data, with errors concentrated in the general class and densely packed regions.

Synthetic Data Generation

Synthetic data generation creates artificial examples, simulations, and datasets to train, test, or protect machine-learning systems.

See all 625 papers

Grounded and Faithful P&ID Reasoning: Constraining Vision-Language Models with Recovered Evidence Graphs
Sep 5, 2026Prathamesh Gadekar, Sagar Srinivas Sakhinana, Venkataramana RunkanaVLM Reasoning Faithfulness in VLMs
Tata Research Development and Design Center, Pune, India 411057
GraphRAG for Engineering Diagrams: ChatP&ID Enables LLM Interaction with P&IDs
Mar 23, 2026Achmad Anggawirya Alimin, Artur M. SchweidtmannGraph Retrieval-Augmented Generation Agentic RAG
Process Intelligence Research Group, Department of Chemical Engineering, Delft University of Technology
LLMs in Process Diagram Engineering: From Optimal PFDs to Validated P&IDs
Jul 21, 2026Timur Zakarin, Sergei Voitov, Sergei Shumilin +1Large Language Model-Guided Optimization Workflow Automation
Skoltech, AI Center, Moscow, Russia · AIRI, Moscow, Russia

Sep 5, 2026J/K move · Enter open · S save

Grounded and Faithful P&ID Reasoning: Constraining Vision-Language Models with Recovered Evidence Graphs

Prathamesh Gadekar, Sagar Srinivas Sakhinana, Venkataramana Runkana

Tata Research Development and Design Center, Pune, India 411057

VLM Reasoning Faithfulness in VLMs Visual Question Answering Tool Use in VLMs

Open paper AskPDF arXiv

Piping and Instrumentation Diagrams (P&IDs) are the authoritative maps of process plants: isolation, maintenance, and HAZOP decisions depend on what connects to what. Vision-language models describe these sheets fluently, yet they often invent or miss process connections---and an invented or missed link can reverse an isolation or reachability call, so a plant decision cannot trust a fluent answer that was never checked against the linework. We instead recover an explicit graph of the drawing---its symbols, the process connections between them, and the tags that name them---and then require the model to answer only by querying that graph through seven read-only operators, so a topology claim is returned only when it cites the query results that support it. On TopoPID-VQA, a new suite of 3000 topology questions over these sheets, Graph-Grounded Harness (Ours) raises exact match accuracy from 36.7--41.3% under image-only prompting to 74.3--76.0% for Qwen3-VL-4B, Qwen3-VL-8B, and Gemma-4-E4B. It does so on an imperfect substrate: on Digitize-PID dataset the recovered graph scores F1 0.742 on exact process connections, and 0.801 once symbols and tags are pooled in. The residual errors track that gap---grounding pays off where the recovered graph is right, and perception error still breaks topology questions where it is not.

Abstract

Synthetic Data Generation

Explore similar work

Grounded and Faithful P&ID Reasoning: Constraining Vision-Language Models with Recovered Evidence Graphs

GraphRAG for Engineering Diagrams: ChatP&ID Enables LLM Interaction with P&IDs

LLMs in Process Diagram Engineering: From Optimal PFDs to Validated P&IDs