
What sensors are actually equipped before a robot gets called 'autonomous'
"Autonomous" on a spec sheet is a claim about software sitting on top of a specific, checkable parts list, and the list is shorter than the word implies on more than one AGIBOT platform.
- LiDAR and RGB-D are the Ultra-tier option on two AGIBOT lines, not the standard one.
- The X2 Basic ships with neither, by AGIBOT's own listing; the X2 Ultra adds 3D LiDAR, an RGB-D camera, front and rear cameras and touch sensors to the same body. The base A3 ships with two RGB cameras, GPS and UWB positioning and no LiDAR or RGB-D at all; the A3 Ultra adds LiDAR, RGB-D and GPS-RTK to that same camera pair. Buying "the A3" or "the X2" without checking which trim is on the order is the most common way a spec sheet's autonomy claim and the crate that actually arrives disagree.
- The two platforms with no stripped-down variant carry the fullest suites as standard.
- The A2 Ultra is sold in one configuration, and it is the richest sensor line AGIBOT publishes: LiDAR, RGB-D, a fisheye camera, stereo cameras, a microphone array, force and torque sensing and an IMU. The G2 is the other single-configuration platform, and its own listing is dual LiDAR, RGB-D, RGB cameras, an 8-channel microphone array, ultrasonic sensing and fingertip visuo-tactile sensing. Neither is sold as a bare version, so the sensor question for these two is answered before the order is placed rather than at the trim level.
- The quadrupeds' camera and LiDAR specs trade off against each other, and the two D1 models make different trades.
- The D1 Max carries twin 192-line LiDAR reaching 120 m across a 360-by-90-degree field, alongside GNSS-RTK, ultrasonic sensing to 5 m and four auxiliary lights, plus two FPV cameras at 8 megapixels with a comparatively narrow field of view. The D1 MaxPro halves the LiDAR line count to a single 96-line unit but states an accuracy figure the D1 Max's own listing does not, 1.5 cm, and swaps the D1 Max's narrower cameras for two wide-angle 2-megapixel units instead. Neither spec sheet is simply better than the other: one buys resolution and range certainty, the other buys a wider field of view and a stated accuracy number, and which trade matters depends on whether the job is spotting a small defect or covering a wide approach.
- LiDAR and depth sensing are what let a platform find something nobody showed it, not what let it move.
- A platform with neither, like the X2 Basic, still executes a taught sequence precisely and repeatably at a fixed position; what it does not do is locate a part whose position varies or a route nobody mapped by hand. That is the actual line the sensor list draws, and it is a cheaper question to ask before an order than after a robot arrives unable to do the one thing a proposal assumed it could.
Ask for the sensor line item by item, by platform and by trim, before "autonomous" or "AI-powered" gets treated as a specification on its own. On this range it names a parts list, and the list is not the same on every SKU that shares a platform's name.




