Mahlis product
Atari Eyes
A non-visual navigation research concept exploring how computer vision and understandable sensory cues may support safer movement.
Current status: Research and early prototype design
Origin of Atari Eyes
The Atari Eyes name emerged in 2024 while exploring computer vision and ways a machine might translate visual surroundings into information a person can understand without relying on a screen.
The central question is not simply whether a camera can identify an object. The harder question is whether a system can communicate useful spatial information quickly, consistently and without overwhelming the person using it.
The navigation problem
A navigation aid must help a person understand what matters now: whether a path is open, where an obstacle is located, whether something is moving and whether the available information is reliable enough to support the next action.
Excessive alerts can become distracting. Unclear alerts can be dangerous. A useful system therefore needs to prioritize information and communicate uncertainty instead of pretending that every observation is correct.
Research direction
Environment sensing
Explore camera and distance-sensing methods for identifying nearby obstacles, pathways and movement.
Non-visual communication
Compare audio, vibration and other understandable cues without assuming one method works for every person.
Priority and timing
Communicate urgent nearby information before less important descriptions of the wider environment.
Confidence boundaries
Make uncertainty visible or audible so the prototype does not present an unreliable observation as a fact.
First prototype questions
- Can the prototype identify a clear path and a nearby obstacle?
- Can it distinguish stationary objects from relevant movement?
- Which cue patterns are understandable without training overload?
- How quickly can information be communicated after detection?
- How does the system behave in darkness, glare, rain or visual clutter?
- How clearly does it communicate low confidence or sensor failure?
Human participation
Non-visual navigation technology should not be designed around assumptions made without the people expected to use it. Prototype evaluation must include appropriate participation from blind and low-vision people, accessibility specialists and orientation and mobility professionals.
Participants should understand what is being tested, what information is collected and what the prototype cannot safely do.
Safety boundaries
Atari Eyes is not currently a mobility aid approved for independent navigation. It must not be presented as a replacement for a white cane, guide dog, orientation and mobility training or human judgment.
Early tests should occur in controlled environments with a defined safety plan. A prototype failure must not place a participant in traffic, near an unprotected drop or in another uncontrolled hazard.
What comes next
- Define a single controlled navigation scenario.
- Document sensor and communication assumptions.
- Build a limited test rig for nearby obstacle detection.
- Measure delay, false detections and missed detections.
- Review findings with appropriate participants and specialists.