LABS
APPLIED RESEARCH EXPERIMENTS
RESEARCHING INTELLIGENCE
Osinto Labs is where we conduct applied research in a commercial setting.
We learn best by doing and so dedicate time to building and breaking things - testing ideas and hypotheses in the real world, pitting them against the cold reality of physics and commerce.
We're inspired by what's been achieved at outfits like Alphabet's X and Lockheed Martin's Skunk Works and seek to build something similarly ambitious and disruptive.
We list on this page a few of the experiments we're working on in the hope of finding collaborators and backers. Partners on a project might come with cash, equipment, expertise, people - or a combination of them all. We both work on discrete projects keeping findings private and sometimes 'build in public' - sharing findings and data for broader benefit.
If you see something that's of interest and you want to work on it with us - simply get in touch.
We don't come with preconceptions - instead focusing on getting things done. Experiments can be spun out as separate ventures, they might equally die a death after a matter of weeks if hypotheses don't stand the test of real world scrutiny.
We're not precious about ideas - so come with criticism, tell us why we're crazy or ill-informed - we learn from it all!
DEFINING THE UNKNOWN
STRUCTURE AMIDST CHAOS
Working at the forefront of anything can be chaotic. Experiment cards help us bring structure and discipline to our research.
The Experiment Card below is Version 0.1 - the intent is to provide a guiding framework that helps us gather thoughts, direct efforts and lay out a succinct vision of what we're trying to do, why and what we need to make it happen.
These are experiments - so we expect many / most of them to fail, and that's fine. Though of course we're always aiming for success!
The intention is to get started, quickly, and sometimes to finish an experiment just as fast. If a hypothesis is quickly disproven, we simply stop and move on.
EXPERIMENT TITLE
EXPERIMENT SUBTITLE
Description
Explain in a few sentences what this experiment seeks to achieve, and why.
What we want to do
Outline the work the Osinto team intends to undertake within the framework of this experiment and the anticipated outcomes through which we'll use to measure success.
Our unfair advantages
We like to play in areas where we bring some unique insights or leverage - they're outlined here.
What we need:
This is where we list what we're lacking in order to get started, it'll nearly always consist of:
-
Money - a budget to cover experiment costs
-
Non-financial resources - this might include specialist equipment we need for a project
-
People - who do we need to make this happen? Defining expertise we don't have, but that the experiment requires
EXPERIMENTS
EXPERIMENT I
BIOACOUSTIC MONITORING PROTOTYPE
NETWORK OF LOW COST AUDIO SENSORS COUPLED WITH MACHINE LEARNING MODEL(S) TO AID IN PERSISTENT DETECTION AND IDENTIFICATION OF WILDLIFE SPECIES FOR BIODIVERSITY MONITORING
Description
We hypothesise that a confluence of low-cost consumer electronics (eg. USB microphones, Raspberry Pi computers, smartphones, lithium-ion batteries, small solar PV panels) together with open source Machine Learning (ML) models, make it possible to build an array of sensors for the persistent, automated detection and monitoring of wildlife, at significantly lower cost than has been possible to date using more manual, sporadic methods of measurement (eg. on-site surveys by ecologists).
With an increasing number of 'rewilding' and biodiversity monitoring projects ongoing worldwide, and an increasing range of attractive subsidy mechanisms coming into force to encourage such projects, we understand there to also be a growing need to monitor species present at a given location over time, so as to prove net gains in biodiversity eg. in order to receive subsidy payments, as well as to aid in wider species monitoring and conservation efforts.
At present much of this work is done manually by ecologists, but only for the limited time periods they are able to be on-site eg. with photographic or sound recording equipment, during which there is also inevitably some disturbance of an ecosystem through human presence.
An automated, persistent bioacoustic monitoring and identification system should aim to reduce cost (vs. repeated site visits) and dramatically increase data volume and quality. We also hope to prove that use of machine learning models to sort signal from noise, de-duplicate and pre-categorise data can aid human ecologist in their analysis of monitoring data.
What we want to do
Establishment of a prototype device and small test network in a biodiverse area in a national park in the UK to see if the hypotheses above stand up to practical experimentation.
Anticipate a 3-month project from outset.
Outcomes expected to include:
-
Determination of viability of producing such a network
-
Determination of forecast cost to provide bioacoustic-detection-network-as-a-service
-
Feedback from prospective clients for a network on functionality and pricing
Our unfair advantages
-
24/7 access to private plot of land in South Downs National Park with 50+ bird and mammal species already identified
-
Access to pool of professional ecologists on weekly basis for advice and feedback
-
Initial research completed to identify possible ML libraries, possible device design and component list, competing technologies at mature end (eg. radar based systems for bird detection at offshore wind farms) and lower end (smartphone apps, commercial audio capture devices used by ecologists in UK)
-
Prospective reserved access to a cluster of low cost, specialised compute in the UK for ML model training
-
Direct professional links to one of UK's foremost rewilding projects for feedback and possible later, commercial scale test
What we need
-
£50,000 - MLOps Engineer time, Data Scientist time, hardware procurement, cloud services setup for data storage, budget for GPU-accelerated compute rental for ML model training
-
We have all expertise and contacts within our existing network to complete this project
OSINTO - RESEARCHING INTELLIGENCE
