About TokyoScale
Utility metering, IoT, and manufacturing—cloud platforms and practical AI built into operator dashboards.
TokyoScale is headquartered in Tokyo (802, 5-21-2 Matsue, Edogawa-ku, Tokyo 132-0025, Japan). We work with water, gas, and electric utilities, manufacturing plants, and infrastructure operators—designing and building from meters and sensors to cloud dashboards as one product.
What we believe
Telemetry earns trust when operators use it—losses become visible, recovery speeds up, and decisions follow one thread. More dashboards are not the goal; defined abnormal, who acts, and when recovery happened are.
We are not an AI billboard agency. We deliver proven metering hardware (Ningbo Water Meter, Shandong Rongxian Instrument, Ningbo Xingyuan Meter, Yuhuan Sierjia Valve), cloud infrastructure (MQTT, time-series stores, multi-tenant UIs), and practical AI (leak, abnormal usage, and equipment fault patterns)—always paired with operational discipline.
How we work
Engagements vary, but the backbone is consistent: listen before tooling, sequence before big-bang rollouts, steward after launch.
1. Discover and measure
We observe how work happens in practice—where data lives, night-shift handoffs, spreadsheets that hold half the truth. The output is shared clarity: priorities, quick wins, and longer bets flagged honestly as conditional.
2. Phased rollout
Sequence matters:
- Stabilize and instrument — truth before guesses
- Remove repetitive failures — small papercuts erode morale
- Scale what keeps trust — compound where adoption is proven
Device registries, MQTT connectivity, dashboards, anomaly detection—not automation for its own sake.
3. Adoption alongside software
We embed with your teams, not only a distant service desk. Escalation drills, night-shift-friendly alerts, documentation lean enough that people actually use it.
4. Operate and improve
Vendors ship updates, regulations change, teams turn over. Stewardship—not handoff and abandonment—is the default.
Principles
Escalation before thresholds — define abnormal and notify paths before drawing charts.
Respect operator know-how — seasoned plant and field staff carry scarce insight.
Explainable AI — ML only where signals and labels are honest; human handoff wired into the UI.
Next step
Contact us—even for an early feasibility conversation before paths harden.