Measurement systems moved into the forest, mound and museum object
A field rig assigning individual capuchin tasks, resistivity scans reading a mound and a sensorized amphora fill all place measurement beside the subject itself.
Science··Morning
A cognition experiment recognizes individuals in the forest
The CapuchinAI system from Emory University and the Georgia Institute of Technology brings controlled laboratory-style cognition tasks to white-faced capuchins living at the Taboga Forest Reserve in Costa Rica. A wooden housing about 51 centimetres tall combines a camera, touchscreen, Raspberry Pi, rotary-motor food dispenser and 3D-printed parts with a battery pack that runs for eight hours. The face-recognition model was trained on images of six individuals and identified those six capuchins with 97 percent accuracy across still images, video and live footage. When the system recognizes an animal, it presents a task suited to that individual's stage and automatically releases a reward after the correct touch; an unfamiliar animal receives a basic habituation stimulus. A total of 16 monkeys interacted with the prototype during the pilot. The researchers are developing tasks for learning, impulse control, cognitive flexibility and short- and long-term memory, and that task set remains under development. The reported output demonstrates a method that closes the loop between identity, stimulus, response and reward in the field.[1]
An unexcavated mound was read with electrical current
Electrical-resistivity mapping at the Coëby cemetery in Brittany examines the interior of a protected archaeological structure before full excavation. The elongated earthen mounds TRED30 and TRED31 are between fifty and eighty metres long, stand up to one point eight metres high and have not been completely excavated. In 2020 researchers inserted small electrodes into the soil and passed a safe current through it; because soil conducts more easily than stone, changes in resistance traced different materials below the surface. Resistance increases measured near the centres of both mounds resembled the profile of granite blocks previously measured at nearby dolmens. In TRED30 the signal strengthened with depth, while TRED31 showed a stone structure closer to the surface. The team says the clusters might be rough human-shaped standing stones like those associated with Passy-type monuments. Resistivity does not directly show the layout or purpose of the stones, and the researchers say complete excavation is needed for confirmation. The survey therefore narrows where and what to look for rather than replacing excavation.[2]
A repair piece became a continuing measurement point
Koç University's amphora method places the measurement device inside a missing portion of a historical object. Hundreds of overlapping photographs become a three-dimensional model through photogrammetry; the absent geometry is reconstructed digitally and a matching fill is produced by additive manufacturing. Sensors for temperature, humidity, pH and physical stress are embedded in that fill. The method was tried on late-Byzantine Günsenin IV amphorae at the Bandırma Archaeology Museum and on the collection of the Minori Villa Romana e Antiquarium in Italy. Initial museum measurements recorded daily changes in temperature and humidity together with force patterns from a stress sensor. Because these variables contribute to salt migration, deformation and microfractures, the fill does more than complete the missing form: it follows conditions affecting the object itself. The forest rig carries a behavioural task to the animal, the resistivity survey brings measurement to the protected mound surface, and the sensorized fill places monitoring in the repair piece. In each case the measurement loop is built where the subject remains instead of moving the subject into a laboratory; the reported scope remains a pilot, a survey and a method evaluation.[3], [1], [2]
Related columns
For more information on this topic, you can read the related columns.
