MOVA AstraX and TactiPercept: arms and touch at IFA 2026

A mower that tries to remove an obstacle instead of driving around it, and a cleaner that responds to physical contact: MOVA's IFA 2026 showcase moves beyond the suction-power race. AstraX and TactiPercept suggest an interesting direction, but a technology presentation is not a substitute for testing at home.

Checked: 4 September 2026. Analysis of MOVA's materials, not a RoboMorrow hands-on review. Lead image: an AstraX visualization published by the manufacturer.

AstraX: an arm intended to clear the way

MOVA describes AstraX as a mower with an arm and interchangeable tools for moving objects, collecting low-positioned fruit and watering. The announcement reviewed here provides neither a price nor a sales date for this model. We treat it as a technology showcase, not a ready-to-buy recommendation.

The practical question is whether owners can stop preparing the lawn before mowing. Avoiding an object and safely picking it up are different tasks. The first requires finding a route; the second also requires deciding whether the object should be moved, how to grasp it and where to put it down.

Not every obstruction should be removed from its current position. A toy, a cable, irrigation equipment and a living animal require very different responses. We would therefore judge the arm on more than an impressive grasp. Correctly refusing to act in an ambiguous situation matters just as much.

TactiPercept: contact becomes another input

The separate TactiPercept technology uses 128 pressure sensors, according to MOVA, to supplement visual information with touch. It is not the garden-arm project. Nor should its announcement be interpreted as a promised software update for robots already sold.

For an owner, the useful outcome would be gentler interaction and fewer rescues, not a particular sensor count. Contact with a heavy table leg is different from contact with a lightweight ornament. Evaluation needs varied objects, surfaces and arrangements, rather than one demonstration beside a prepared wall.

What would we check before recommending it?

Our questions for these technologies are straightforward, but demanding:

• Risk recognition: when does the robot refuse to move something, and how does it notify the owner?
• Repeatability: does the task still succeed after changes to object position, lighting and ground conditions?
• Placement: can a moved object end up blocking a path, damaging a flowerbed or obstructing another cleaning route?
• Failed grasps: can the device safely abandon an unsuccessful attempt?
• Maintenance: how much time is needed to clean the mechanism and, where applicable, change tools?

Useful robotics means completing the entire task. If every grasp requires somebody to adjust the setup and restart the machine, the labor-saving case weakens. Conversely, there is no need to demand human versatility. A clearly defined, dependable task can be worth more than an expansive list of loosely demonstrated abilities.

An object-handling arm is not an edge trimmer

IFA is bringing ideas for both moving objects and finishing lawn edges. Those address different buying needs. If a strip of uncut grass is your main frustration, ask about the result at the boundary, not simply whether a specification mentions an arm.

For a purchase this season, start with our robot mower buying guide. A trade-show newcomer belongs in a buying comparison once the exact commercial offer, conditions and service arrangements are clear. A visualization is not a basis for inventing a price, coverage figure or warranty.

The right goal: less work before the robot starts

MOVA's most interesting proposition is reducing the preparation people do before automation can begin. Verification must include failures and exceptions, not just successful demonstrations. For now, this is a direction to follow rather than a reason to replace a working device.

For another ambitious design shift, read about Roomba Duo and its onboard companion. Our IFA 2026 Live Desk tracks further announcements, pricing and availability.