Whales and dolphins in managed care or research settings
Intermittent-reinforcement fish dispenser
A feeder that pays out fish on a variable schedule, so a cetacean learns that its own signals and choices can make something happen — the first step of any two-way exchange.
- Variable-ratio and variable-interval reinforcement schedules, tuned across trials to keep engagement high without over-feeding.
- Randomised payouts (one, two or three fish) to sustain interest.
- A simple control interface so trainers and researchers set the rules and see every payout.
What it builds on: Variable reinforcement schedules are among the most studied results in behavioural science (Ferster & Skinner, 1957), and positive-reinforcement training is standard practice for marine mammals in managed care.
Partners we need: Aquariums and marine research centres with cetacean-care teams; hardware and food-safety engineers.
Honey bee colonies
Beekeeper hive link
Let a beekeeper “read” what the colony is saying and, over time, answer: decode waggle dances into a map of where foragers are going, and listen to the hive for signs of stress or swarming.
- A camera on the observation frame decodes waggle-run angle and duration into direction and distance.
- Hive sound and vibration monitoring for colony state.
- Longer term: answering back with a robotic dancer, building on published robot-bee work.
What it builds on: The waggle dance is the one animal signal decoded to the point of meaning (von Frisch, 1967); automated dance decoding and robotic dancers have been demonstrated in research (e.g. Landgraf et al., 2018).
Partners we need: Beekeepers and apiary associations; bee researchers; makers of hive sensors.
Dolphins and whales
Cetacean two-choice playback game
Play recorded clicks or whistles and let the animal choose between two targets. Its choices show which signals it tells apart — a way of asking the animal, not guessing.
- An underwater speaker plays sperm-whale codas or dolphin whistles mixed with background noise.
- The animal picks target A or B; a reward follows a correct choice.
- Signal-detection statistics (d′) measure how well each signal is discriminated, against chance.
What it builds on: Two-alternative forced-choice and playback designs are standard in animal cognition and bioacoustics research.
Partners we need: Marine mammal cognition labs; aquariums; underwater acoustics engineers.
Dolphins
Echolocation feedback game
Turn a dolphin’s own echolocation into the controller: its click trains aimed at targets trigger responses, so it can play by “looking” with sound.
- High-rate hydrophones capture click trains with precise timing.
- Click trains aimed at a target are classified and trigger feedback (light, sound or reward).
What it builds on: Dolphin echolocation and its use in target discrimination tasks are well documented in the bioacoustics literature.
Partners we need: Dolphin research programmes; hydrophone and signal-processing specialists.
Dogs and cats at home
Button-board telemetry for dogs and cats
Pet owners already use recordable word buttons with their animals. Add careful logging so button presses can be studied alongside the animal’s voice and body language.
- Log sequences of button presses (for example “outside + play + now”) with timestamps.
- Record vocalisations and movement alongside, so presses can be checked against context rather than taken at face value.
What it builds on: Whether button presses carry meaning is an open research question; the value here is collecting clean, analysable data rather than claiming the animals are talking.
Partners we need: Companion-animal cognition researchers; pet-owner communities; button-board makers.