Analytics

Concentration, failure modes and trends computed from the sourced incident log. Regenerated with every refresh.

Findings

• 3473 documented incidents across 206 of 2148 tracked companies; 330 entries adversarially source-verified; 86 rest on regulator records.
• av trucking carries the largest share of the log (132 entries); other leads on serious events (S4+: 40).
• 59 entries involve a fatality or catastrophic loss; legacy train-control failures and industrial robot cells account for the largest clusters.
• Fires account for 8 of 16 energy storage entries - the category's dominant failure mode.
• Cyber entries (20) concentrate in consumer robotics and Chinese-manufactured platforms, mostly disclosed-and-patched (S1-S2).
• 2025 logged 111 entries vs 88 so far in 2026; regulator aggregate records (NHTSA SGO, FDA MAUDE) are dated to their retrieval month and inflate the current year.

Where serious events concentrate

Entries by category and severity. S4+ = serious injury, total loss, grounding, fatality.

Dominant failure modes by category

Share of each category's record by incident type (percent of category entries).

Volume and severity over time

Category concentration and trend

The largest holder's share indicates whether a category's record is systemic or dominated by one operator.

Index trajectories — top companies

RobotRisk index recomputed as of each year end (decay re-anchored), so the line shows how each record accumulated and fades.

Highest RobotRisk index

Fatality and catastrophic-loss ledger

Every S5 or fatality entry on file, most recent first.

Evidence quality

Every entry carries a source; confidence grades the corroboration, source class records whether a regulator document backs it.

Compare companies