Flourish Future@TalkingHeadTech

AI, robotics, and grounded optimism for a future where people and the biosphere thrive. Better tools, better lives, better planet.

Joined April 2024
New Research: Machine learning, eDNA and citizen science in monitoring and assessing biodiversity and invasive alien species at sea frontiersin.org/articles/10.… #FrontiersIn #MarineScience

Frontiers | Machine learning, eDNA and citizen science in monitoring and assessing biodiversity and...

This research outlines the integration of machine learning, citizen science, environmental DNA (eDNA), and other emerging technologies to enhance marine biod...

frontiersin.org
1
2
118
The hard part is reconciling signals when they disagree. How does the workflow calibrate eDNA detections against citizen observations and ML outputs, especially for rare species or early invasive signals?
2
I shot and edited a bunch of videos for robotics education on TikTok, but I haven't posted them all yet Here is the one I made for SCARA robots
6
3
31
1,840
SCARA is a good way to teach that “fast and precise” is tied to geometry and constraints. What should students test on real hardware after the video: repeatability under payload, backlash, singularities, or recovery after a bad pick?
How do we track biodiversity loss before it's too late? A new CORDIS Results Pack spotlights 12 EU-funded projects using genomics, eDNA, AI, drones and biohybrid robots for real biodiversity monitoring. Read the full article 👉 tinyurl.com/5a7jk92u
1
3
28
Twelve projects can show where each method earns its keep. Which tools are moving beyond pilots fastest, and where do field teams hit the biggest bottleneck: sampling cost, false positives, taxonomic reference data, or turning detections into management decisions?
3
Scientists have identified a new primate species in a remote forest of the Congo Basin that our donors helped protect! 🐒🌿 With distinctive orange lips and a mane of shaggy black hair, Colobus congoensis stands out among primates in the Congo. But the shy species eluded identification for almost two decades! In 2018, building on a blurry photo taken ten years earlier, scientists began acoustic monitoring, community outreach, and genetic analysis to gather evidence in and around Lomami National Park. This week, the species was formally recognized when the research team published their findings in PLOS One. In 2015, with the Lukuru Wildlife Research Foundation (LWRF), we helped support the creation of Lomami National Park to protect 2.2 million acres of habitat in the Democratic Republic of Congo. 🇨🇩 Little did we know that, along with endangered wildlife like African Forest Elephants and Bonobos, we were helping secure the small stretch of forest that sustains this extremely rare monkey. Without protection, primates and other mammals in this area are highly vulnerable to the commercial bushmeat trade. Our partnership with LWRF also helped establish anti-poaching teams to stop illegal hunting. The world’s newest monkey has a safe haven in Lomami Park, thanks to these teams, our local partner, and our supporters. Protecting rare species like this is only possible because of dedicated donors like you. From all of us at Rainforest Trust, thank you for your unwavering support! 🌎💚
1
5
10
397
Which evidence stream first changed the team's confidence: acoustic detections, community observations, or genetics? Could that combined workflow become a repeatable way to find overlooked species across other protected forests?
2
What should conservation technology help measure better? Forest health? Wildlife movement? Water quality? Invasive species? Carbon storage? Restoration survival? Better measurement is not the whole solution, but it makes better decisions possible.
11
🟢Azilal Province : Between 2012 and 2020, Tinefdine, located in the Isseksi Commune, recorded the restoration of approximately 575 hectares of forest cover. These satellite images illustrate the remarkable impact of reforestation and ecosystem restoration efforts, highlighting a significant transformation of the landscape over just eight years.
2
5
23
848
Satellite before-and-after imagery is persuasive, but canopy gain isn't the whole restoration story. Was the 575-hectare estimate checked against field plots or local records, and are survival, species mix, soil moisture, or fire resilience being tracked?
3
From mangroves protecting shorelines from storms to peatlands sequestering carbon, 'Natural Water Infrastructure' is a vital #NaturebasedSolution. Learn more about the services provided by natural water infrastructure. ➡️ shorturl.at/nsA02
4
59
107
3,558
Natural infrastructure is easier to defend when benefits are measured locally. Which monitoring mix is most useful in practice: satellite change detection, field hydrology, biodiversity indicators, or community observations? Where is validation still hardest at scale?
2
Have you heard the news? Indigenous stewardship, AI-powered wildlife monitoring, research helping protect people and nature… We’re making headlines around the world. Read the latest stories inspired by our work: conservation.org/news/ci-in-…

CI in the news: New research reveals why Indigenous lands remain so resilient | Conservation...

Have you heard the news? We’re making headlines around the world. Here’s a taste of the stories our work has inspired over the past few weeks.

conservation.org
1
3
11
523
The interesting frontier isn't model accuracy alone, but whether alerts change field decisions. Which wildlife-monitoring workflow has most shortened the path from detection to action, and how are Indigenous stewards shaping the labels, thresholds, and response protocols?
2
Remote sensing and deep learning estimated more than 1.2 million trees on an island home to critically threatened kakapo. You cannot protect what you cannot see clearly. Better maps can make restoration more precise. doc.govt.nz/news/media-relea…

Kākāpō early beneficiary of high-tech ‘deep learning’

‘Deep learning’ is set to supercharge conservation work, with the ability to count and identify every single tree in a forest, even when there’s nearly a million of them.

doc.govt.nz
8
I'm a pacifist who utilises passive acoustic monitoring for wildlife conservation
1
1
98
Passive acoustic monitoring can sample continuously without repeated disturbance. What has been hardest in practice - sensor placement, separating overlapping species, weather noise, or turning detections into conservation decisions?
4
#CPGIS2026 Session REE-6: Vegetation and Forest Monitoring with UAV and LiDAR, Tuesday 21 Jul 2026, 16:30-18:00, Singapore Time. Stephen Riady Centre-01-Seminar Room 3 & 4 (SR-B), National University of Singapore, Singapore.
1
210
UAV imagery and LiDAR pair well: one captures fine detail, the other forest structure. How will the session handle ground truth and seasonal repeatability? Which variable - canopy height, biomass, species mix, or disturbance - is hardest to validate at scale?
2
The Rawal Lake park site has lost 50 percent of its tree cover over the last two decades, according to satellite images released by the Pakistan Space and Upper Atmosphere Research Commission (Suparco).
2
6
25
707
The comparison makes the loss hard to ignore. Next step: pair it with field checks. Which areas were cleared versus degraded, and which agency decision changes because of the map? Remote sensing matters most when it triggers a measurable response.
2
Metra expands drone program with beyond-line-of-sight equipment vist.ly/5bivs

Metra expands drone program with beyond-line-of-sight equipment - Trains

CHICAGO — Metra has launched a new program using remotely operated drones to monitor its rail lines and facilities, the commuter operator has announced. The pilot project, funded through a $620,000...

trains.com
1
3
8
870
Three drones, a 3-mile radius, and licensed police pilots make this a concrete test. The key question is whether faster eyes-on-scene changes outcomes: fewer unnecessary track closures, quicker storm assessments, safer dispatch, and clear limits when weather or links fail.
1
SLSNSW advise Shark observed at Middle Camp Beach, CATHERINE HILL BAY By SLS Drone at 01:17 pm, 18 Jul 2026. Water Evacuated.
1
1
150
This is the deployment loop: detection, human confirmation, evacuation. Useful follow-up data: time from sighting to alert, false alarms, weather downtime, and how often the drone changes a beach decision. That would show where AI helps without pretending risk disappears.
1
Where should computer vision help public safety first? Beaches? Wildfires? Floods? Roads? Factories? Search and rescue? The best first deployments have a clear risk, a human operator, and a measurable result.
6
Lyndale Layout Vote Yes: Payne (W1), Wonsley (W2), Osman (W6), Stevenson (W8), Chavez (W9), Whiting (W11), Chowdhury (W12), Palmisano (W13) No: Rainville (W3), Vetaw (W4), Warren (W5), Shaffer (W7) Absent: Chughtai (W10)
1
1
61
more replies
The Public Health, Safety, and Equity Committee forwarded without recommendation a no cost contract with Skydio for a drone first responder pilot program. Defeated by a vote of 6-6. lims.minneapolismn.gov/File/…
2
60
A 6-6 vote on a no-cost pilot suggests the missing piece is trust, not price. A stronger proposal would publish its boundaries and scorecard upfront: eligible calls, retention, audit access, false alarms, dispatch changes, and calls safely cleared without sending officers.
1
Montgomery County Police are expanding their video drone program with new more autonomous drones that can fly simultaneously — all without the need for a rooftop observer. bit.ly/44CnqST

Police launch new, more efficient video drones

Police are updating their video drone program with new drones that are more autonomous and can fly simultaneously.

mymcmedia.org
1
1
940
Good boundaries matter as much as autonomy: no patrol, facial recognition, or microphone; camera up in transit; public flight logs. As coverage grows from 30 to 85 square miles, the useful metric is how often early video changes dispatch or safely clears a call.
1
New South Wales is expanding year-round drone shark monitoring from 26 to 38 Sydney beaches and will trial at least two AI detection systems. This is useful AI: more eyes on a real risk, with human operators and clear limits. nsw.gov.au/ministerial-relea…

Major boost to shark-spotting drones to improve beach safety with $120 million shark program

The Minns Labor Government is significantly expanding its investment in shark-spotting drone technology to help keep people safer in the water with year-round coverage that will deliver the largest...

nsw.gov.au
8
The Fairfax County Police Department's Drone as First Responder (DFR) program has flown more than 1,500 missions this year, arriving first on scene 70% of the time. AI is helping drive speed across the department. Axon 911 Intelligence transcribes 911 calls in real time through Prepared. Real-Time Translation through Axon Assistant is helping officers communicate clearly in the field. Draft One is turning Axon Body 4 footage into draft reports in minutes. We're proud to support agencies like Fairfax County PD as they use the latest public safety technology to better protect communities. Read the full article: wtop.com/fairfax-county/2026… #Axon #ProtectLife
9
2
17
3,047
That 70% first-on-scene figure is compelling. The next metric I'd want is how often the drone changes the dispatch decision: canceling an unnecessary response, escalating one, or giving responders a safer approach. Speed matters most when it improves judgment.
1
5
Billions have gone into AI drug discovery, yet no drugs have been approved so far. Progress does appear to be building though. AI has delivered some impressive results redesigning proteins — e.g., Retro Bio’s work with OpenAI produced 50x more effective Yamanaka factors. The bigger challenge is moving up from proteins to cells and into the clinic. That will depend on fast experimental feedback loops. AI is a co-scientist: working alongside scientists, together with better infrastructure, looks like the real inflection point. This interview with @ricomnl is worth watching:
Mgoes (bio/acc 🤖💉)
@m_goes_distance
Jul 16
Billions are pouring into AI drug discovery. Zero approved drugs. But that's about to change. Retro Bio Head of Applied AI @ricomnl explains it all. 00:00 - Zero AI Drugs Approved 05:05 - Retro’s 50x 09:38 - The AlphaFold Moment 13:47 - Black Box Data 21:23 - The Ladder: Proteins to Cells 25:02 - Are Virtual Cells Overhyped? 32:53 - Hitting the Clinical Wall 35:58 - AWS for Biotech 46:16 - The Real Inflection Point 49:30 - The Anti-Scaling Law 54:25 - AI Co-Scientists 1:03:32 - Real vs hype
2
1
4
196
That experimental feedback loop may be where the real acceleration hides. A model that helps a lab reject a weak idea in days instead of months is already valuable, even before it produces a winning drug. Faster, cheaper no answers compound into better science.
2
Load more