Design Digital Twin Platform - System Design | WinJob

Digital twin platform framing and industrial asset ontology

Design a multi-tenant platform that mirrors turbines, production lines, and building systems as software twins fed by OT telemetry, BIM/CAD geometry, and operator workflows.

Numbers to state early

Metric A: 500k twins
Metric B: p99 sync 2s
Metric C: DTDL models

Mechanism

Each twin is a versioned graph node with typed relationships (contains, feeds, drives) and property bags for setpoints versus live readings. Ingest normalizes heterogeneous protocols into canonical observation events keyed by tenant_id:asset_id.

Failure and edge cases

Clock skew across PLCs can reorder events; use ingest-time watermarks plus per-asset sequence numbers before updating twin properties. Never apply stale telemetry over newer operator overrides.

When discussing Digital twin platform framing and industrial asset ontology, anchor on 500k twins and p99 sync 2s before naming managed services. Explain how Physical Asset → Twin Graph → Telemetry ordering survives retries and regulator audits.

Java

javaOne Dark Pro

1 public final class TwinRevision {
2   private final long revision;
3   private final Instant observedAt;
4  
5   public TwinRevision(long revision, Instant observedAt) {
6     if (revision < 0) throw new IllegalArgumentException("revision");
7     this.revision = revision;
8     this.observedAt = observedAt;
9   }
10  
11   public boolean isNewerThan(TwinRevision other) {
12     return other == null || this.revision > other.revision;
13   }
14 }

Python

pythonOne Dark Pro

1 from dataclasses import dataclass
2 from datetime import datetime
3  
4 @dataclass(frozen=True)
5 class TwinRevision:
6     revision: int
7     observed_at: datetime
8  
9     def is_newer_than(self, other: "TwinRevision | None") -> bool:
10         if self.revision < 0:
11             raise ValueError("revision")
12         return other is None or self.revision > other.revision

TypeScript

typescriptOne Dark Pro

1 export interface TwinRevision {
2   revision: number;
3   observedAt: string;
4 }
5  
6 export function isNewerThan(a: TwinRevision, b: TwinRevision | null): boolean {
7   if (a.revision < 0) throw new Error("revision");
8   return b === null || a.revision > b.revision;
9 }

Why interviewers care

Digital Twin Platform interviews reward crisp scope, explicit trade-offs, and failure stories—not generic microservice diagrams.

Interview checkpoint

Name one failure story for Digital twin platform framing and industrial asset ontology that proves you understand real outages, not happy-path diagrams.

Key Highlights

•500k twins
•Physical Asset → Twin Graph → Telemetry
•Design a multi-tenant platform that mirrors turbines, production lines, and building systems as software twins fed by OT.

1	public final class TwinRevision {
2	private final long revision;
3	private final Instant observedAt;
4
5	public TwinRevision(long revision, Instant observedAt) {
6	if (revision < 0) throw new IllegalArgumentException("revision");
7	this.revision = revision;
8	this.observedAt = observedAt;
9	}
10
11	public boolean isNewerThan(TwinRevision other) {
12	return other == null \|\| this.revision > other.revision;
13	}
14	}

1	from dataclasses import dataclass
2	from datetime import datetime
3
4	@dataclass(frozen=True)
5	class TwinRevision:
6	revision: int
7	observed_at: datetime
8
9	def is_newer_than(self, other: "TwinRevision \| None") -> bool:
10	if self.revision < 0:
11	raise ValueError("revision")
12	return other is None or self.revision > other.revision

1	export interface TwinRevision {
2	revision: number;
3	observedAt: string;
4	}
5
6	export function isNewerThan(a: TwinRevision, b: TwinRevision \| null): boolean {
7	if (a.revision < 0) throw new Error("revision");
8	return b === null \|\| a.revision > b.revision;
9	}