See the interfaces among data, control, hardware and people.
Robotics technology stack · Layer 03
Manipulation and dexterity
From grasping and contact sensing to compliant hands, bimanual coordination and recovery when real objects behave unexpectedly.
Evidence statusMature for structured tasks; research frontier for open-world dexterityLast reviewed: 11 August 2026
03
FUURAA thesis
Dexterity is a closed-loop systems property: object coverage, sensing, mechanics, planning and recovery matter more than a visually impressive hand motion.
Reading method
Turn a technology label into an engineering chain, testable claims and explicit boundaries.
Translate capability into task, latency, failure and recovery.
Separate standards, independent measurement, research and first-party claims.
State what cannot be inferred from a demo, benchmark or interface.
System breakdown
Four interdependent layers determine whether the technology can enter real work.
Each layer shows its role and the failure signal most worth watching.
End effector
Fingers, suction, tooling and compliance define the contact modes and objects the robot can handle.
Contact sensing
Force, torque and tactile signals reveal slip, seating, insertion and excessive contact.
Planning and control
Collision-aware motion, impedance and force control turn a target grasp into stable physical interaction.
Task recovery
Regrasping, clearing jams and asking for assistance determine useful autonomy over long runs.
Engineering evaluation
Five checks turn abstract capability into reviewable system evidence.
Record normal performance, failure, recovery and human cost—not only the best-looking result.
- 01
Use representative object sets
Include transparent, reflective, dark, deformable, damaged and closely packed objects when relevant.
- 02
Score the whole cycle
Measure approach, grasp, transfer, placement, verification and recovery—not pick success alone.
- 03
Measure damage and near misses
Record drops, crushing, scratches, collisions and unsafe force, including successful-looking cycles.
- 04
Vary presentation
Test clutter, pose, fill level, packaging and upstream variability across real shifts.
- 05
Report intervention burden
Include tooling changes, teaching, jam clearance and cleaning in throughput and cost.
Scope boundaries
State what the evidence supports—and what it does not.
- 01
A hand with many degrees of freedom is not necessarily more robust or economical for a bounded task.
- 02
Lab object sets often underrepresent wear, contamination, packaging variation and upstream process errors.
- 03
Human-like motion should not be mistaken for human-level tactile reasoning or adaptability.
Sources and evidence status
Keep the source, date, evidence identity and reading boundary visible.
This page prioritises standards bodies, public measurement programmes, official project documentation and original research disclosures.
01
NIST · reviewed 2024-06-24
↗02Robotics Test Facility
Describes task-based artifacts and measurement methods for mobility, sensing and manipulation under controlled challenges.
ISO · 1998-04
↗ISO 9283 — Performance criteria and related test methods
Provides established performance criteria for manipulating industrial robots.