Sony's robotics work connects machines that move with systems that sense, learn and communicate. AIBO and aibo made that combination familiar through a companion animal; SDR and QRIO explored humanoid movement; precision robotics applies related engineering strengths to very different tasks.
The most useful way to read this history is to ask what each project was built to investigate. A consumer companion, a laboratory platform and a medical research prototype have different goals and different evidence of success. The overview below was reviewed on October 3, 2026.
A selective map of Sony robotics
| Milestone | Main contribution | Project status |
|---|---|---|
| 1999: original AIBO | An autonomous entertainment robot designed around continuing interaction. | Historical consumer product family. |
| 2002: SDR-4X | Small biped platform combining motion control, perception and communication. | Research prototype. |
| 2003: QRIO | Humanoid platform demonstrating integrated walking, running and jumping control. | Historical research and demonstration platform. |
| 2018: aibo returns | A new companion design with enhanced AI capabilities. | Consumer product generation; check regional service and sales information. |
| 2024: microsurgery assistance | Precise instrument control and automatic instrument exchange. | Development prototype in the cited announcement. |
| 2026: university aibo collaboration | Research-oriented hardware and beta development tools for partner laboratories. | Announced research collaboration. |
Sony's robotics milestones and 25-year aibo retrospective supply the historical framework. The original AIBO line was discontinued in 2006; aibo returned in 2018. That interruption is part of the story, rather than an unbroken sequence of annual product releases.
AIBO and aibo: interaction as the task
A companion robot has to make movement meaningful to a person. The timing of a head turn, a response to touch and the relationship between sound and posture affect how an interaction feels. Perception, behavior selection and mechanical execution therefore become parts of the same design problem.
A useful distinction is between a robot recognizing an input and a person interpreting its response. A movement can communicate playfulness to an owner without establishing that the machine experiences that feeling. Judge the interaction by what it supports: engagement, understandable responses and a relationship a person chooses to sustain.
Sony's account of its AI robotics evolution places AIBO within a wider history of combining AI, sensing and mechatronics. That connection explains why companion robotics is also a research setting: real homes and repeated interactions introduce variation that a single stage demonstration cannot capture.
When considering a consumer aibo, use Sony's aibo FAQ to check the model, region, companion app and cloud-plan requirements. An owner's experience depends on that service environment as well as the physical robot.
SDR-4X and QRIO: making a body respond
The March 2002 SDR-4X announcement described a small biped with 38 joints, stereo cameras and multiple microphones. Sony combined posture control with perception and communication, allowing the prototype to respond to changes in its surroundings. The significance lies in coordinating those abilities in one moving machine.
Sony's December 2003 QRIO announcement then described integrated control for walking, jumping and running. Moving from walking to running changes the balance problem because the robot can briefly have both feet off the ground. A successful demonstration establishes a specific motion under particular conditions; broader usefulness depends on reliability, recovery and the task the motion serves.
This is a useful lens for watching humanoid videos today. Look for what the robot senses during execution, what was prepared in advance and how it responds when the situation changes. The same distinction appears in robot dance and creative performance.
Precision robotics: a different measure of success
Sony's May 2024 microsurgery assistance announcement described a prototype that transfers an operator's hand movements to miniature instruments and can exchange instruments automatically. The research combines sensitive controls, small mechanisms and detailed imaging.
The reported experiment involved animal blood vessels. At the time of that release, Sony explicitly described the system as being in development and lacking the relevant Japanese medical-device approval. Treat that source as evidence of a research milestone; it does not establish current clinical availability or patient outcomes.
The broader engineering lesson is that “better movement” depends on purpose. A companion needs expressive timing; a walking robot needs stable whole-body motion; an instrument assistant needs controlled movement at a much smaller scale. One spectacular capability is an incomplete measure of another application.
The 2026 aibo research collaboration
On July 17, 2026, Sony announced collaboration with laboratories at the University of Tokyo and UC Berkeley. It described supplying research-use aibo prototypes and beta development tools. The partners identified interests including social interaction, shared experiences and learning through social and tactile interaction.
This offers a concrete direction for physical AI: learning systems that encounter the world through a body. The announcement establishes the collaboration and its intended platform. Published experiments and evaluated results will be the evidence for what researchers achieve with it; the announcement itself is not a retail product launch.
Follow the research by its evidence
For any new Sony robot story, identify three things: the task, the development stage and the reported result. Read a dated release for what was announced, a technical paper for the experiment and limitations, and current product documentation for what an owner can actually use.
That approach also makes the history easier to appreciate. Sony's contribution spans interaction and control, while the wider robotics timeline adds industrial work, mobile reasoning and planetary exploration. Together they show why progress in robotics takes several forms.