
TECHNICAL BRIEFS
HOW WAS IT ENGINEERED?
Concise engineering intelligence on consequential robotics work: mechanisms, control, numbers, limitations, and links back to original sources when available.
8 briefs
A robot dog went into Sellafield. The interesting part was the swab.
A controlled trial with blue chalk reveals the work between sending a robot into a restricted area and trusting the sample it brings back.
The gearbox does not create torque for free
A robot joint needs torque, speed and manageable inertia at the same time. The ratio on the gearbox is only the opening bid.
Perseverance is slow on purpose. Its wheels have other priorities.
Six driven wheels, passive suspension and a carefully limited driving envelope show how Mars mobility is negotiated between mechanics, terrain and software.
Why an underwater robot might want a very long leash
Jason and Nereid Under Ice use very different cables. Comparing them reveals a design choice about energy, communication and mechanical disturbance.
A nuclear robot needs an exit strategy
Fukushima Daiichi’s retrieval records show why camera replacement, controlled withdrawal and sample containment belong in the original robot design.
The suction cup has not read your AI benchmark
Reliable picking begins at the contact patch. Suction physics and Berkeley’s Dex-Net research explain why recognizing an object is only part of grasping it.
The inspection robot remembers the route. Does it know where it is?
Spot’s GraphNav documentation reveals the practical work behind repeatable inspection: map references, recognizable features and a surprisingly consequential printer setting.
Two confident sensors can still get a robot lost
Sensor fusion needs meaningful uncertainty, consistent frames and genuinely distinct information. Adding inputs alone does not supply any of them.
