Workspace
Trends & forecasts
Model failure risk, battery fade, and workload so teams can plan ahead.
Refresh: Daily 06:12 UTC · Default planning horizon
High-risk devices
441
Model score
Forecast horizon
30D
Default
Model refresh
Daily
Auto
Accuracy
0.81
AUC proxy
High-risk devices — ranked
AI-assisted operational insightModel-ranked failure risk (0–1); filter to focus triage.
- 0.91Critical
Battery: −4.2% cap / mo
- 0.84Elevated
Battery: −2.8% / mo
- 0.79Elevated
- 0.72Elevated
- 0.68Elevated
Battery: −1.1% / mo
- 0.61Watch
- 0.58Watch
- 0.55Watch
- 0.52Watch
Battery: String 2 aging
- 0.48Watch
Predicted corrective events
AI-assisted operational insightPoisson–gamma blend on telemetry features — 90% posterior bands.
Hazard index · 30d resolution
Uptick mid-series reflects scheduled vendor windows and census overlay — not a model fault.
Battery fade — fleet cohorts
AI-assisted operational insightRolling 90d regression on effective capacity; used as a prior in device risk scores.
- Telemetry patches · ICU−2.1% / mo · health 78%
2023 cell batch — plan swap Q3
- Infusion fleet (wireless)−2.8% / mo · health 71%
Correlated with high-traffic pods
- Portable patient monitors−1.3% / mo · health 84%
Within expected decay
- Temp probes · cold chain−3.6% / mo · health 62%
Probe flex fatigue cluster
Workload forecast
AI-assisted operational insightPredicted corrective labor for 30d vs staffed capacity — peak week Apr 7–13.
Baseline trend 380h — add 36h vs last horizon for same cohort.
Overtime risk
Medium
Contractor lift
+32h