JupiteX Get the app
Science & Technology22 Nov 2026 · about 7 min

This featherweight, flying robot carries its own sensors and computers

The brief

The robot’s onboard sensing and computation mean it carries the electronics needed to gather information about its motion and process that information while flying. Earlier versions had to stay connected to a lab motion-capture system for navigation. Adding these capabilities is important because it could let tiny flying robots work in places where that external system is unavailable. The electronics package weighs 244 milligrams. It includes an inertial measurement unit for estimating the robot’s attitude, an optical flow sensor for estimating lateral speed and position, and a time-of-flight sensor for measuring height. A microcontroller runs the computations. Using these onboard tools, the robot followed a complex trajectory and landed precisely on a sunflower outside the motion-capture environment. This is a demonstration of sensory and computational autonomy, not full independence. The robot still receives power from an offboard source, and it cannot determine its absolute position, so it may drift off course. The researchers also designed a fast algorithm to estimate orientation, angular velocity, and position. The work points toward possible uses such as crop monitoring, pollination, and industrial surveillance, but the article describes those as future potential.

01

What does it mean for this microrobot to have onboard sensing and computation?

The robot’s onboard sensing and computation mean it carries the electronics needed to gather information about its motion and process that information while flying. Earlier versions had to stay connected to a lab motion-capture system for navigation. Adding these capabilities is important because it could let tiny flying robots work in places where that external system is unavailable.

The electronics package weighs 244 milligrams. It includes an inertial measurement unit for estimating the robot’s attitude, an optical flow sensor for estimating lateral speed and position, and a time-of-flight sensor for measuring height. A microcontroller runs the computations. Using these onboard tools, the robot followed a complex trajectory and landed precisely on a sunflower outside the motion-capture environment.

This is a demonstration of sensory and computational autonomy, not full independence. The robot still receives power from an offboard source, and it cannot determine its absolute position, so it may drift off course. The researchers also designed a fast algorithm to estimate orientation, angular velocity, and position. The work points toward possible uses such as crop monitoring, pollination, and industrial surveillance, but the article describes those as future potential.

02

How small is the robot, and how much do its onboard electronics weigh?

This flapping-wing microrobot is roughly the same size and shape as a microcassette, and it weighs less than a paperclip. That tiny scale is central to its promise: the researchers point to tasks where a larger robot cannot be used, including pollination and industrial surveillance. But its small size also makes it challenging to add the sensors and computing needed to navigate on its own.

The onboard electronics package weighs 244 milligrams. It holds the sensors and microcontroller for sensing and computation. The sensor suite includes an inertial measurement unit, an optical flow sensor, and a time-of-flight sensor. To make room for this payload without losing flight performance, the researchers increased the size of the robot’s soft artificial-muscle actuators and tuned their geometry.

The electronics now allow the microrobot to estimate motion without lab motion-capture cameras, and it demonstrated a precise sunflower landing. The robot is not yet fully untethered: it remains connected to an offboard power source. The research shows how a very small flying robot can carry navigation electronics, while leaving further development necessary before the suggested field uses become practical.

03

How does the robot estimate its movement and position without motion-capture cameras?

The microrobot estimates its movement using a small onboard sensor suite and computation, rather than depending on an offboard motion-capture system. That matters because the earlier robots had to remain connected to the lab system to navigate. The new approach lets the robot sense and compute while flying outside that environment, as shown by its trajectory and sunflower landing.

An inertial measurement unit estimates attitude, an optical flow sensor estimates lateral flight speed and position, and a time-of-flight sensor measures height. The researchers also developed an algorithm that rapidly estimates orientation, angular velocity, and position while simplifying calculations to reduce computational cost. They found that continuous, faster position estimates performed better than slower, more accurate estimates.

The sensors do not give the robot an absolute position. Without that reference, it can drift off course, even though it can estimate its movement. The article also notes that the demonstration still uses an offboard power source. The onboard navigation is therefore an important step, while the robot’s position limits and tethered power remain part of its current reality.

04

How did the researchers increase the robot’s ability to lift the added electronics?

The researchers needed the microrobot to carry sensors and computing without sacrificing the speed, maneuverability, and collision robustness of its earlier design. They changed the soft actuators, also called artificial muscles, that flap its four wings. The challenge was to add lifting capacity without reducing the system’s output power or harming flight performance.

They carefully tuned the actuators’ radius and length, increasing their size by 50 percent. The adjustment boosted the robot’s net lift forces by 46 percent. The researchers say they tuned the robot’s geometry so lift increased proportionally with the additional actuator weight. The resulting lift-to-weight ratio was 1.9, enough to carry the electronics package, which weighs 244 milligrams.

The added lifting ability made onboard sensing and computation possible, and the robot demonstrated a precise landing outside the motion-capture environment. However, the demonstration still relies on an offboard power source. The work shows one way to increase payload while preserving flight performance, but the article presents uses such as pollination and crop monitoring as future possibilities rather than established field applications.

05

What could microrobots do for crops, such as pollinating blossoms or spotting pests?

The article describes possible agricultural uses for microrobots that can fly close to plants. They could help with precision agriculture in regions where natural pollinator populations have been decimated by habitat loss. Their small size may also make them suitable for tasks where a larger robot cannot be used, such as moving among crops without damaging them.

For example, flying robots could weave through almond trees or blueberry bushes and carefully pollinate delicate blossoms. They could also hover near plants to monitor crop health and spot pests before those pests become widespread. The demonstrated robot’s onboard sensors estimate attitude, lateral flight speed and position, and height, while its onboard computation supports navigation.

These crop tasks are examples of future potential, not results of the reported demonstration. The researchers showed the microrobot following a complex trajectory and landing on a sunflower outside the motion-capture environment. It still draws power from an offboard source, and it cannot determine its absolute position, so it can drift. The article says this technology could be promising if the research direction succeeds in the future.

06

What important limitation remains, even though the robot can sense and compute onboard?

The clearest remaining limitation is power: the microrobot is still connected to an offboard power source. Its demonstration shows onboard sensing and computation, but not a fully self-powered flying robot. This matters for proposed uses in fields or other settings, because the article presents the work as a step toward future applications rather than a finished agricultural tool.

There is also a navigation limitation. The onboard inertial measurement unit, optical flow sensor, and time-of-flight sensor estimate attitude, lateral movement and position, and height. But without the precision of the attached motion-capture system, the robot does not know its absolute position. The article says this can cause it to drift off course. The researchers developed a fast algorithm to estimate orientation, angular velocity, and position.

Even with those limits, the robot followed a complex trajectory and landed precisely on a sunflower outside the motion-capture environment. The demonstration establishes that it can sense and compute autonomously, as the researchers describe it. The article identifies pollination, crop monitoring, and industrial surveillance as possible future tasks, while noting that the robot remains powered offboard.

07

What is a lift-to-weight ratio, and why does it determine whether a flying robot can carry a payload?

A lift-to-weight ratio compares a flying robot’s lifting force with its weight. It matters because the robot must generate enough lift to support itself and any payload it carries. If added electronics increase weight without enough added lift, the robot may not be able to fly as intended. The article uses this measure to explain why the actuator changes mattered.

The researchers report a lift-to-weight ratio of 1.9 after increasing the soft actuators’ size by 50 percent and boosting net lift forces by 46 percent. The added lift allowed the tiny robot to carry an onboard electronics package weighing 244 milligrams. That package contains sensors and a microcontroller for sensing and computation.

The ratio helped make the onboard electronics possible, and the robot demonstrated a precise landing on a sunflower outside the motion-capture environment. It is still connected to an offboard power source, however. The article presents the work as a milestone with potential for future tasks such as pollination and crop monitoring, not as proof that those applications are already in use.

This brief was written by AI from the original reporting and checked by other models. Names, figures and quotes come from the source; read it for full context.

Read more in the JupiteX app

Pulse is free. New stories every 4 hours, each one broken into the questions that explain it.

Or read more news on the web