It has come from nowhere. Probably the best kept secret in the entire tech industry (note I say Tech industry and not medtech.) A hyper stealth project between consumer products giants - AI chip makers - AI software companies. But not surprisingly none of the medical device companies.The project was only a rumour until now that went by the name of “Project Dark Star”. But last week I was taken to their vast and super secretive headquarters and given exclusive access to Unreel Robotics amazing and mind blowing new robot called “Quasar.”
I’m embargoed to say which colossal company are behind this - but they have more money than they will ever need - with one source saying they have infused $30 Billion into this project. And everyone of those companies is a name you would know and say “I wondered when they would enter this market.”
Their aim is simple says CEO Anna Wang - “We want to deliver the future of surgery now. Humanity cannot wait another few decades.”
So what is Quasar and why am I reeling in shock?
Simply put - this robot can perform any - and I mean any surgical procedure a human can do today. Open - Lap - Robotic - endovascular - endolumenal - see n treat to full resection. Also percutaneous energy delivery for zero incision cellular destruction of small tumours etc.
It is primarily aimed at soft tissue - but later this year the hard tissue tool sets and navigation software will be revealed.
Let me run down the amazing feature set and then we will break down each one individually:
Fully autonomous. And I mean autonomous. It brings itself into the OR, it configures itself for the right procedure based on pre-operative insights. It selects the right procedure itself - open - lap / robotic or interventional and then performs the entire procedure from start to finish under total self autonomous control and decision making. (I will explain fully the software behind this called Clinical Think.)
The robot can form a tower configuration like a boom, or it will set itself up as a modular system - or a combination system. It even has micro robot deployable modules.
Console… what console? There is no pilot for this thing.
It will apply classical rigid multi arm robotics - or open surgical robotics - or flexible endolumenal - or rendezvous - all coming from the same arms - but with different final joints that hot swap as per patient needs.
It has amazing open and lap vision systems called “Fusion Sight” which takes multimodal imaging in combination and fuses the different data sources to be able to see through organs in real time.
It has force feed back and “Reel Feel Haptics” that replicate inputs as if a surgeon was actually feeling the tissue. Including sensing if it is warm, wet, pulsatile. Quasar feels like a surgeon.
It deploys next generation advanced energy, PFA, ultrasonics, microwave and a new technology for tissue welding that eliminates the need for stapling or other mechanical anastomosis technologies.
There’s more… but let me dive in.
Full autonomy means full autonomy
The Clinical Think software is the brainchild of David Parry - a name not familiar to medical devices - but well known in the world of autonomous self driving cars.
“Imagine taking the world’s best clinical insights - fusing them. Then creating a vast artificial data set based on that knowledge. Billions and billions of procedures simulated every day,” says David. “You let our AI simulate a million years of surgery and learn from it. It understands everything pre-operative and creates a tailored plan for that specific patient.”
Clinical Think draws on all of the possible pre-operative data available.. Even requesting more pre-clinical work up to optimise the operation. And if it can’t get it - it fills in the gaps itself from it’s almost unlimited data library using AI.
“The software takes all of that data and then acts as the most knowledgeable clinician on the planet. It doesn’t think like a surgeon or a cardiologist or a radiologist - it thinks like all of them combined all at once and has an unbiased way of thinking. This clinical fusion thinking then determines the best procedure for the patient. Including no operation or intervention - wait and see is often the right answer. It determines if it is a biopsy with see n treat. If a minimally invasive surgery is best - and what surgery. It determines if an endolumenal approach works better or a non invasive energy therapy.”
Clinical Think never stops and never stops learning. While you are reading this article Clinical Think is busy doing millions of virtual patient care pathways and treatments - learning by the second. Getting better by the minute.
“The key is the feedback loop,” says David. “Humans are notoriously bad at being creatures of habit. It is hard to get a human to immediately change thinking and behaviour. Clinical Think is not like that. It uses data - simulated data and real world evidence as a feedback loop. Using a global access into all electronic health records we can see pre treatment - treatment and outcomes. Clinical Think is taking millions of outcomes a day and combining it with billions of synthetic data points and fusing that knowledge in real time to make small and large treatment adjustments. And that knowledge flows out instantaneously to the entire fleet of Quasar robots. It does’t wait five years for the results of a limited RCT and then take another five years to put that into practice. We are talking hours. It’s instantaneous knowledge dissemination and behavioural change.”
So this holistic feedback loop of thinking is a revolution in medicine. But Clinical Think does not stop there. This is where the science fiction kicks into high gear - because now it will act autonomously.
Clinical Think sets out the best patient pathway - discusses this using a natural language model with the patient - as a type of clinical chat bot. Which by the way was utterly amazing and I got used to it way faster than I thought for my fictitious hernia procedure. The amazing thing was the chat bot seemed to understand my level of knowledge by my responses and was listening to my desires - my fears - what I wanted from my treatment. It also nuanced that I was a physiologist that had been in surgery a long time because it wanted to know more about me. So it adapted its language to speak in terms I would understand. I did this twice - once as “me” and once as “someone that understands nothing about healthcare.” The Clinical Think software nuanced the conversation and consent to ensure I fully understood in layman’s terms what it was suggesting. I was amazingly comfortable with the interaction. It then explained the full process - guided me on what I needed to do to maximise my pre op work up - and then gave me a bio sensor it wanted me to wear until after the post op period. I won’t cover that today - but that biosensor is a key part of work up and follow up and perioperative data collection.
This thing just does procedures to perfection
When it comes to the operation / intervention, Clinical Think is thinking like a holistic surgeon - OR nurse - OR manager and even anaesthetist. Yes - it guides the anaesthetist on the best protocol to use for this procedure to attain the best clinical result. It guides them pre - peri and pos op via an app.
Clinical Think has also scheduled all of the ORs in the right order to maximise efficiency - and sends scheduling to all departments to ensure patients - anaesthesiologists and nurses all arrive in the right place at the right time to maximise workflow. Each staff member has a Clinical Think app that they can use and follow. But also say “I’m sick today” and Clinical Think will sort the staffing in real time on the fly to ensure cases work is not disrupted. If not it re-arranges the OR scheduling to make it work better for that staffing team.
It manages inventory - streamlines OR consumables and small (think of them as Amazon robots) set up the OR equipment needs ready for Quasar.
Now there’s too much to talk about today to cover how Clinical Think manages everything about the OR - from set up, to operative notes, to billing and payment collection. That’s for another day.
Quasar as a surgeon - oh my God. It's a God.
Clinical Think has sent all the information it needs with the plan to the robot modules of Quasar. I saw multiple set ups. It self selects its modules and those modules navigate the OR (with no cables by the way.) The self driving modules are small - about the size of a human. (Sorry I cannot show images yet.)
They then position themselves around the bed. And this is where it gets interesting.
They stay as modules if that is considered the best set up and approach to the patient.
Or they assemble themselves autonomously into one large boom tower. It’s like watching a transformer. It then (from a small circulating accessory module) selects the end joint for each robot arm that will carry either open instruments - or robotic wristed instruments - or percutaneous energy devices - or even endolumenal devices. Quasar prepares itself for that patient’s tailored procedure. Equipping itself with the right toolset.
Clinical staff (for now) just prep the patient on the table. And then Quasar takes over. In lap robotic procedures it already knows the full body shape and size of the patient - and even works out their abdominal shape when insufflated. 3D lidar systems on Quasar are analysing the entire OR in real time including the patient. It places the trocars in the “right place” as per best practice - but it also knows where every vessel and nerve in the abdominal wall are… and adjusts placement to avoid those structures. We will get into the fact it can see what surgeon’s cannot.
The speed it works at is incredibly efficient - and trocars are in and insufflation (tailored to the patient) is set. In this robotic case it selected the instrument end effectors autonomously - self loaded them - self inserted them and with zero hesitation it performed a hernia, or a gall bladder, or a colon resection - in incredibly fast times. Why? Because it doesn’t do any inefficient steps or manoeuvres. It doesn’t fumble with needles or spend time adjusting them in the needle holder. It’s a super computer that doesn’t miss. Ever. Think of electronics assembly robots - bang bang bang. And it has simulated that actual procedure in its digital twin 1000 X before the patient hitst the table. It has nuanced the procedure for optimising time and outcomes.
To see it place a running suture line is like watching a sewing machine - fast accurate - perfect.
Knot tying is interesting because it has worked out that surgeon knots are not that good or efficient and it has already self taught itself a new knot it calls K32.
Understand there is no human intervention - because the would just slow things down. The human is the weakest link in the OR chain.
So how is it doing this magic?
The first secret to this they say is the “Fusion Sight.”
Unreel Robotics know that white light spectrum that clinicians use is super limited. And Quasar is not limited to human vision. In fact - I can’t say too much - but you wouldn’t recognise the vision system for lap work or even open work. It’s nothing like a laparoscope you’ve ever seen. I can’t even call it a laparoscope.
But it works by having access to all kinds of imaging modalities. Mutispectral - hyperspectral - NIR - microscopic - Lidar - Laser speckle - and some new modalities that can detect perfusion and tumours via novel imaging - even at depth. Qualitatively and quantitatively. It "sees" tumour boundaries at the cellular level. Well it has already seen them a thousand times in its digital twin based on the prior MRI.
It has all of that pre operative imaging data and can fuse them together with real world physics using landmarks as fiducials - and in real time create 3D mesh versions of all the organs - fully deformable in real time with super high accuracy.
It then creates a second digital twin in real time with all the sub structures and Quasar can basically see though organs as it it is not limited to the human visual spectrum or way we perceive vision. It is not just seeing the structures in the abdomen. It is seeing the fused structures in its computer (brain) and pairing the two or three models.
Interesting side note: it has a Quasar MRI pneumatic arm set that can work inside an operative MRI in real time - and of course if C-arms are being used for anything (more hard tissue later) there is no operator to worry about for X-rays. Same goes for interventional procedures. Talk about protecting staff.
So combining all of this vision - physics modelling - planning - it then adds on a layer of AI based structure recognition - so it is seeing the whole surgical scene in real time as structures - “That it understands”. It knows where a vessel is - it knows what it is - it knows what its tissue dynamics “should be” it can see pulse - oxygen saturation and even has knowledge of the root of that vessel deep in tissue. And it has that in real world fused vision in front of it in the patient - and in its digital bio model in its “head” - work that out.
But then Quasar goes one step further forwards as it does not rely on vision alone. Quasar is a sensory haptic robot. I’m not talking about just gross force feedback like we see today - I’m talking about full on haptics. It has sensors in the instrument tips that give it what it calls “True Touch Haptics”
It feels if the tissue is wet, elastic, warm or cold, rough or smooth, hard or soft. It is as if it is a surgeon with their hands inside the patient palpating tissue. And it understands all of that with forces in real time and processes it 100,000 X faster than a human.
With the combination of Fusion Sight and True Touch - the machine, at lightning speed, can fully understand - and that is how it dissects planes to perfection in second… and almost bloodless. It places sutures in the right places - as if it was a suturing machine - tying bespoke knots.
I watched over thirty surgical video procedures and then a live telepresence view as a Quasar performed an autonomous Whipples in under 45 minutes.
I can only describe watching a procedure as if watching the world’s best surgeon at 5X speeded up (but it’s real time.). It is surreal - and it just never wastes a single movement.
The instruments are sublime - with fully wristed end effectors and most of them multifunctional and nearly all can deliver an array of energy. Each instrument will work at a perfect level - and then the sensors in the instrument will suggest changing it. It’s not time or lives based - it’s functionality based. Economics is about to change.
Let’s talk energy and “stapling” or not stapling
One thing you will not see with Quasar is monopolar energy - anywhere. Their teams have deemed it archaic and downright “Dangerous!”
Instead they have created what they call “Personal Energy” systems. These are fused energy modalities that understand the tissues and deliver energy based upon the tissue characteristics and the tissue response during energy delivery. Be that ultrasonic, advanced bipolar, microwave and pulsed fields. Often switching between them in microseconds as the tissue changes under energy delivery.
The system uses AI to understand the right delivery and if a different end effector is needed - the system swaps out the instrument via a holster system and reinserts the right tool for the job.
There are no bedside assistants and instrument nurses needed with the Quasar OR system.
It then delivers the right energy that adapts to the tissue. Sealing vessels, opening planes, energy delivered during dissection - all controlled by complex algorithms that use constant feedback loops at the tissue level to ensure minimal thermal spread and collateral damage. -Optimal tissue effect without damaging tissue and with the Pulse Field Energy affecting the metabolic state of the tissue to induce faster healing, downgrading inflammation.
There are no “settings” there are no “40 cut and 40 coag” - this applies blended energy that is tailored to every second of the surgery. To every patient. To every tissue type in real time. For Quasar there is no limit to data processing. It doesn't get cognitive overload.
Four instrument arms working simultaneously is quite normal for complex procedures.
And let’s talk stapling. Or rather let’s hear what the head of Tissue Interfaces Anika Bakshi has to say. “It is still unbelievable to me that in 2025 we are still holding tissue together with archaic metal staples. A constant guessing game of tissue thickness and effect. At Unreel we don’t believe in leaving unnecessary things behind in the body.”
Unreel has worked with two of the world’s leading institutions in tissue properties. And they have created “Tissue Fusion”.
Using a proprietary energy technology that acts at the sub cellular level - when two tissue structures are placed together - and the energy applied the tissues - they fuse into one another as a viable living matrix. At the cellular level they are literally one tissue. It is as if the tissue had grown that way. Their rapid histology analysis system built into Fusion sight allowed me to see in real time that effect in an live anastomosis.
“This was one of the hardest challenges,” Anika says. “But we used AI to help us design the right energy modality and instrument design. We were quite surprised with what the AI came up with. But having AI that understands proteins at the molecular level and how energy can unwrap them and then combine them was the breakthrough. We are still not quite sure how it does it. But it works. Staples are history.”
My question was how sure are they that it works as well as staples. And the answer came with their constant leak detection system. An evolution of the insufflation system (a system that constantly monitors the physiological condition of the abdomen and applies physiologic vapour, muscle relaxants and anaesthesia agents - and also monitors for leaks.
It combines that with what I will call “The eye in the sky” system. Which is that every port has. Small image sensor that when the images are stitched together it gives Quasar a “view” of the entire abdomen - a bit like those cameras on a car that make it look as if the car is being seen by a virtual camera above.
In effect the eye in the sky is doing the same - looking for leaks, bleeds, any errors, macro changes in the abdomen (or chest). It does the same for open surgery but in higher definition. Think of it as Quasar checking itself and the patient every second of the procedure. It will then use AI to act accordingly. Leaks (if they ever occur) are detected and identified in real time. The AI fixes them.
Erm - so where is the surgeon? Where's the cardiologist?
I thought at some point you might ask.
Well where they are not… is in the OR at the table getting long term ergonomic injuries.
Mark Vantes - Chief of colorectal surgery explains it well. “No human can do surgery like this. It is the combination of all the clinical knowledge and hand eye coordination of best practices from every corner of the world. Stimulated in synthetic data and applied with the up to the second AI enhanced clinical insights.”
He smirks. “Quasar is the best surgeon everywhere at once. It’s the best cardiologist everywhere at once. It’s the best Radiologist everywhere at once. And we have to accept that our days as an operator are over. We can’t do better because our human limitations get in the way.”
Me “So are you redundant?”
Mark “No. We will evolve. Firstly, Quasar won’t be everywhere yet so there is a decade of transition. They have most regulatory approvals across the globe already secured. It can be used anywhere tomorrow technically but will take time to roll out. And with a massive multi billion foundation in partnership it will not be limited to just rich first world countries. Unreel Robotics wants this everywhere. But it takes time to build them - ship them and transition.”
Me “What do you mean transition?”
Mark “Quasar will change what we do as surgeons. Our role will be supervision of a cluster of Quasar robots - probably in multiple hospitals. We will be administrating and giving some human oversight. Quasar should be perfect out of the gate - the few billion simulations predict it works. But patients will still want to know that ultimately there will be supervision somewhere. If you look at the Quasar control room that has been deployed in our hospital - it’s like a command centre. We have a rotation of surgeons - cardiologists - interventional neurosurgeons etc etc. We come in, and a bit like air traffic control view the dashboards and the live feeds. We oversee operations.”
Me “Air traffic control works 24/7”
Mark “And so does Quasar. We have a roster now in the hospital and we are doing shifts. The system is working 24/7 and the efficiency has shot up. We can operate like a factory now as Quasar never gets tired. It self monitors and with AI it understands if it needs preventative maintenance. So there is no down time of Quasar. If an arm has a potential issue it rolls itself out of the OR down the corridor to the Quasar bay. Another arm is already on its way. It’s efficiency personified with no human intervention.”
Shocking to some as this may sound. I personally see it as evolution. 24/7 on demand patient care. I saw demonstrations of robotic lap surgery, open surgery, micro surgery, cardiologic interventions, structural heart procedures, advanced endoscopy procedures. It doesn’t matter - Quasar is a clinician that can do it all - autonomously.
Who can afford this?
This is the part that will shock most people. The predicted level of efficiency, the massive reduction in staff, the predicted elimination of clinical errors and Quasar not only pays for itself - it saves the healthcare system billions of $. Well all simulations so far. But the real world data from the 20 pilot sites is rolling in fast. Because of course it collects all of the financial data and presents the hospital with evidence. And with a crash down to ZERO clinical complications in over 10,000 interventions. It's encoraging.
Not only that - as we see the looming population demographic heading our way - we will need all the automation we can get in healthcare. Quasar is one step. Robo nursing, Robo physio, Robo GPs - will not just be a nice to have they will be a must have. There will not physically be enough staff to deal with the boom in octogenarians we are heading towards. A drop in the young workforce will necessitate Quasar. There won’t be enough surgeons to drive conventional robots the way they do today in the near future.
There will not be enough money to allow inefficiencies, and above all, clinical errors to continue - that cost is colossal. AI and robotics will eliminate errors. We just cannot pay for complications any more.
One of the world’s richest foundations has paired up with Unreel and will give billions to accelerate the deployment of Quasar. They have funded that every single major hospital on the planet can have their first Quasar placed at zero cost. The challenges are building and deploying at that scale. But when you have companies that understand consumer products - and consumer scale - the people behind Unreel Robotics think this is no more fantasy that delivering electric cars but at 1/10th the volume. It’s been done. It is not impossible for non medical device companies. And the factories are almost ready. Giga robot factories let's call them. A drone flight over one of them shows the industrial scale they have built.
Quasar is about to upend the entire economics around surgery - and interventions. We cannot limit ourselves to thinking in the old way about the cost of instruments and the cost of the OR and not understanding the macro economics associated to patient episodes. It feels more expensive at the point of care - but what is the massive impact of Quasar if we eliminate surgical errors (financial and human?)
The march to progress and better health cannot be hamstrung by "bean counting" limitations of seeing the bigger picture. Of course with having the dual systems in parallel costs go up initially - but as we transition to the future of autonomy - AI and robotics we then realise the savings and the efficiencies. It will take the kind of bold vision we see at Unreel Robotics.
Have you got it yet?
So how come you have never heard of Unreel Robotics and Quasar?
Now anyone that has arrived this far down and done even a minimal amount of fact checking will have understood that everything above is a fictitious scenario. But a scenario that I hope felt “it could be real.”
And strangely why does it feel it could be real?
You see what I wanted to do today was paint a picture of a few things in the extreme.
Nothing I have said today as individual parts is actually that fictitious.
There are modular systems and boom systems combined - Kangduo.
There are companies disrupting insufflation - Qaelon
Nvidia is ploughing all kinds of resources for advanced vision and edge AI computing.
Creo Medical has kamaptive tissue technology
We have force feedback and many companies are working on genuine haptics
Johns Hopkins is deep into autonomy as one of multiple examples.
AI is everywhere, and nothing I have said today is fantasy - I mean nothing… it’s all in research or early deployment. Consultative chat bots are being deployed across medicine.
Where I see surgeons going to “overseers” - that my friends is a future I firmly believe in.
Self swapping Instruments have been shown by Sony in its microsurgery demo.
Autonomous scope systems are here today. Autonomous surgical steps are arriving. Lupin Dental did six autonomous dental procedures last week.
Andromeda Surgical is doing AI guided autonomous HoLep procedures as you read.
All of those advanced imaging methods are in research today (including eye in the sky). All of them.
Digital twins are used daily, OR oversight and black boxing is nearly here
Self propelled hospital carts are real today. They are around hospitals today.
Tissue welding is a future step in progress - but advanced stapling with RevMedica is on the horizon.
Central lines are being placed autonomously and blood draws done today commercially - autonomously. Trocars is a simple step up.
AITA does automated OR supplied management today.
Moon Surgical has demonstrated OR scheduling and advanced OR Lidar to scene detect.
It’s here.
Caresyntax is doing much of what I said Quasar would combine in OR overview and more.
Etc etc. The list goes on. This is not science fiction.
One issue is that it is fragmented across many research groups and companies today. If it all came under one visionary group (Unreel) - this could be made reality tomorrow. Of course regulatory is a hurdle - but FDA is getting comfy with AI and autonomy. It’s just time.
So if anyone has a spare $30 billion sloshing around - send it my way and I will build Unreel Robotics (clue was in the name for those awake today…)
You can read about many of these technologies in my previous blog posts. Or follow me on linked in where I discuss this live and in posts. Do me one favour - if you even minimally thought this was real - give my post a like, a cryptic comment and share this article: but don’t give the game away in comments please. My ultimate goal is to paint an incredibly amazing vision of what is on the near horizon. I want all companies to see this is coming - surgeons to feel the horizon and prepare - and maybe inspire an Apple - or a Google - or a Sony - or a Space X to jump in. I want to inspire.
Who’s with me? Who wants to be part of Unreel Robotics?
Enjoy the day
I’m in!! 😉😉
Eerie peak into the future of surgery. Disheartening for any surgical trainees