Latest / Let's Talk Risk! Podcast / Deep Dive: Designing Safety Into Autonomous Robotic Devices

Transcript

Deep Dive: Designing Safety Into Autonomous Robotic Devices · Let's Talk Risk! Podcast

• 0:00
  SPEAKER_01: If you look at the the back end of any major hospital laboratory today,
• 0:05
  SPEAKER_01: you are basically looking at a marvel of modern manufacturing.
• 0:09
  SPEAKER_01: I mean,
• 0:09
  SPEAKER_01: you have these massive multimillion dollar tracks using AI and precise robotics
• 0:16
  SPEAKER_01: to analyze clinical chemistry.
• 0:18
  SPEAKER_00: Oh, yeah.
• 0:18
  SPEAKER_00: The analytical phases, it's completely industrialized at this point.
• 0:23
  SPEAKER_00: It operates with just near perfect mechanical consistency.
• 0:26
  SPEAKER_00: You've got analyzers calibrated to like fractions of a microliter.
• 0:29
  SPEAKER_01: Right.
• 0:30
  SPEAKER_01: Running thousands of samples an hour.
• 0:32
  SPEAKER_01: But then you walk out to the waiting room
• 0:34
  SPEAKER_01: and you see how we actually feed those incredible machines.
• 0:37
  SPEAKER_01: And the fundamental process hasn't changed since,
• 0:39
  SPEAKER_01: well, since the vacuum tube was invented like 75 years ago.
• 0:42
  SPEAKER_00: Exactly.
• 0:42
  SPEAKER_00: We are still relying on a human being.
• 0:45
  SPEAKER_00: Using their naked eye and their sense of touch to manually slide a needle into a vein.
• 0:51
  SPEAKER_01: Yeah, it's wild.
• 0:52
  SPEAKER_01: We have this 21st century analytical automation that is just entirely dependent on
• 0:56
  SPEAKER_01: a highly variable manual 19th century collection technique.
• 1:00
  SPEAKER_00: The vulnerability of the modern clinical lab is quite literally concentrated at
• 1:04
  SPEAKER_00: the point of the needle.
• 1:05
  SPEAKER_01: And that paradox, that is the core of our deep dive today.
• 1:09
  SPEAKER_01: So for those of you listening,
• 1:10
  SPEAKER_01: whether you're in clinical ops, R&D,
• 1:12
  SPEAKER_01: quality assurance,
• 1:13
  SPEAKER_01: or regulatory affairs,
• 1:16
  SPEAKER_01: We are unpacking the clinical benefit risk profile and the engineering behind
• 1:20
  SPEAKER_01: a device that is actively rewriting the regulatory landscape.
• 1:24
  SPEAKER_00: Yeah, we are looking at Vitestra's Aleta,
• 1:27
  SPEAKER_00: which is the world's first autonomous robotic phlebotomy device, or ARPD.
• 1:32
  SPEAKER_01: Right.
• 1:32
  SPEAKER_01: And to really understand the magnitude of this,
• 1:35
  SPEAKER_01: we have to look at the systemic healthcare bottleneck that actually forced the FDA
• 1:39
  SPEAKER_01: to create a brand new regulatory framework just to accommodate it.
• 1:43
  SPEAKER_00: Because on August 19,
• 1:45
  SPEAKER_00: 2026, the FDA granted de novo marketing authorization to the Aleta device.
• 1:51
  SPEAKER_00: And, you know, the de novo pathway...
• 1:53
  SPEAKER_01: It's fascinating, right,
• 1:54
  SPEAKER_01: from a regulatory perspective,
• 1:55
  SPEAKER_01: because it's reserved for things that are entirely novel.
• 1:58
  SPEAKER_00: Exactly.
• 1:58
  SPEAKER_00: There was no legally marketed predicate device for this.
• 2:00
  SPEAKER_00: The FDA couldn't just point to a previous machine and say,
• 2:02
  SPEAKER_00: okay, prove you were substantially equivalent to this.
• 2:05
  SPEAKER_00: They had to invent a whole new classification for automated vascular access systems.
• 2:10
  SPEAKER_01: And they didn't take it lightly.
• 2:11
  SPEAKER_01: Right.
• 2:12
  SPEAKER_00: No, not at all.
• 2:13
  SPEAKER_00: The agency laid down really strict special controls.
• 2:16
  SPEAKER_00: So they required comprehensive clinical performance verification,
• 2:21
  SPEAKER_00: rigorous biocompatibility testing of all the fluid path materials,
• 2:25
  SPEAKER_00: and extensive mechanical testing.
• 2:26
  SPEAKER_01: Right.
• 2:27
  SPEAKER_01: They didn't lower the barrier to entry.
• 2:29
  SPEAKER_01: They built this new highly scrutinized pathway because
• 2:33
  SPEAKER_01: the clinical void it fills is just so desperate.
• 2:35
  SPEAKER_01: I mean, phlebotomy is one of the most commonly performed invasive procedures globally.
• 2:40
  SPEAKER_01: But clinical labs are facing a devastating staffing crisis right now.
• 2:44
  SPEAKER_00: The numbers and the sources we're looking at are staggering.
• 2:47
  SPEAKER_00: We're seeing annual phlebotomist turnover rates in U.S. clinical labs of up to 46%.
• 2:51
  SPEAKER_00: Wow, 46%.
• 2:54
  SPEAKER_00: Yeah.
• 2:55
  SPEAKER_00: And when a phlebotomist leaves,
• 2:56
  SPEAKER_00: the data shows it takes an average of seven to 10 months to recruit,
• 3:00
  SPEAKER_00: hire, and train a replacement.
• 3:01
  SPEAKER_01: Which is just a massive operational gap.
• 3:03
  SPEAKER_01: And that cascades through the entire health system.
• 3:05
  SPEAKER_01: Like if the lab can't draw the blood, the analyzers sit idle.
• 3:08
  SPEAKER_01: And if they sit idle,
• 3:09
  SPEAKER_01: emergency department boarding times skyrocket because,
• 3:12
  SPEAKER_01: you know,
• 3:12
  SPEAKER_01: physicians can't discharge or admit patients without those diagnostic results.
• 3:16
  SPEAKER_00: It leads to severe bottlenecks and massive burnout for the personnel
• 3:21
  SPEAKER_00: who are left trying to shoulder the load.
• 3:23
  SPEAKER_01: It's like having a state-of-the-art automated fulfillment center.
• 3:27
  SPEAKER_01: But the loading dock is entirely dependent on one guy carrying boxes by hand, one at a time.
• 3:32
  SPEAKER_00: That's a great way to put it.
• 3:33
  SPEAKER_00: For decades, drawing blood has been a strict one-to-one ratio.
• 3:37
  SPEAKER_00: One healthcare worker, one patient, one manual draw.
• 3:41
  SPEAKER_00: But Aleta completely fractures that constraint.
• 3:43
  SPEAKER_01: Right.
• 3:44
  SPEAKER_01: We're shifting from this artisanal craft to a scalable supervised workflow because
• 3:49
  SPEAKER_01: under this new FDA clearance,
• 3:50
  SPEAKER_01: a single trained clinical supervisor can oversee up
• 3:54
  SPEAKER_01: to three OLED devices simultaneously in an outpatient setting.
• 3:58
  SPEAKER_00: So you're tripling workforce capacity without compromising clinical throughput.
• 4:02
  SPEAKER_00: And that's the operational leap that hospital leadership is analyzing very closely right now.
• 4:07
  SPEAKER_01: But, you know,
• 4:07
  SPEAKER_01: bridging the gap between a regulatory clearance and
• 4:10
  SPEAKER_01: the actual physical engineering requires solving a massive technical challenge.
• 4:14
  SPEAKER_01: My immediate thought just looking at this concept is about the unpredictability
• 4:17
  SPEAKER_01: of human physiology.
• 4:18
  SPEAKER_00: Right.
• 4:19
  SPEAKER_00: The anatomy.
• 4:20
  SPEAKER_01: Yeah.
• 4:20
  SPEAKER_01: How does a machine safely perform an invasive needle stick on a human being
• 4:25
  SPEAKER_01: without direct manual intervention?
• 4:27
  SPEAKER_01: Like, how does it know where to go?
• 4:29
  SPEAKER_00: Well, the architecture relies on a synthesis of multimodal imaging,
• 4:33
  SPEAKER_00: computer vision, and highly constrained robotic motion control.
• 4:37
  SPEAKER_00: So the target acquisition sequence is built on active redundancy.
• 4:41
  SPEAKER_01: Okay, walk us through it.
• 4:42
  SPEAKER_01: A patient puts their arm in the kiosk, and then what?
• 4:45
  SPEAKER_00: The machine initiates an autonomous scanning protocol using near infrared, or NIR, light.
• 4:50
  SPEAKER_01: Right, and near infrared utilizes specific wavelengths of light that are absorbed
• 4:55
  SPEAKER_01: by deoxygenated hemoglobin.
• 4:57
  SPEAKER_01: So that absorption makes the superficial veins appear to Exactly.
• 5:03
  SPEAKER_01: That 2D
• 5:07
  SPEAKER_00: map gives it the initial trajectory,
• 5:08
  SPEAKER_00: but it's insufficient for an invasive procedure on its own.
• 5:12
  SPEAKER_00: So the system immediately combines that visual map with a 3D ultrasound sweep.
• 5:17
  SPEAKER_00: High-frequency sound waves penetrate the tissue to determine the exact depth,
• 5:21
  SPEAKER_00: diameter, and orientation of the venous structures.
• 5:24
  SPEAKER_00: And then it deploys a really critical safety mechanism,
• 5:27
  SPEAKER_00: which is real-time Doppler ultrasound, to analyze the localized fluid dynamics.
• 5:32
  SPEAKER_01: Because hitting a high-pressure artery during a routine blood draw is
• 5:35
  SPEAKER_01: a severe clinical complication.
• 5:38
  SPEAKER_01: The machine has to definitively know the difference between a vein and an artery
• 5:42
  SPEAKER_01: before it even moves the needle.
• 5:43
  SPEAKER_00: Yeah, precisely.
• 5:44
  SPEAKER_00: So the AI analyzes the acoustic Doppler signature of the blood flow.
• 5:49
  SPEAKER_00: Venous flow is steady, it's low pressure, and it returns to the heart.
• 5:53
  SPEAKER_00: Arterial flow is highly pulsatile, it's high pressure, and pushes away from the heart.
• 5:57
  SPEAKER_01: Right, you can actually hear that pulse on a Doppler.
• 5:59
  SPEAKER_00: Yeah, and the Doppler acoustics translate those flow characteristics into
• 6:02
  SPEAKER_00: a mathematical signature.
• 6:04
  SPEAKER_00: If the signature is a pulsatile,
• 6:05
  SPEAKER_00: the AI flags it as an artery and just completely walls off that area from
• 6:10
  SPEAKER_00: the robotic trajectory.
• 6:11
  SPEAKER_01: I want to highlight a piece of practical engineering here that I found so clever
• 6:15
  SPEAKER_01: because it solves a notoriously messy problem.
• 6:19
  SPEAKER_01: Usually to get a clear ultrasound image, you need that non-sterile sticky acoustic gel.
• 6:25
  SPEAKER_00: Right, which you can't use here.
• 6:26
  SPEAKER_01: Exactly.
• 6:27
  SPEAKER_01: So Vitestro engineered a dual-use fluidic solution.
• 6:31
  SPEAKER_01: During the scan, the system continuously sprays 70% alcohol onto the patient's arm.
• 6:36
  SPEAKER_00: Which is brilliant.
• 6:37
  SPEAKER_00: The continuous alcohol spray acts as the immediate clinical disinfectant,
• 6:42
  SPEAKER_00: so it preps the site to standard clinical protocols while simultaneously serving as
• 6:46
  SPEAKER_00: the acoustic coupling medium for the ultrasound probe.
• 6:49
  SPEAKER_01: It entirely eliminates the need for gel,
• 6:51
  SPEAKER_01: keeping the whole process clean, sterile, and automated.
• 6:55
  SPEAKER_01: But, and this is where I have to push back a bit,
• 6:57
  SPEAKER_01: even with perfect sensors and elegant fluid dynamics, humans are unpredictable.
• 7:02
  SPEAKER_01: Someone comes in dehydrated,
• 7:03
  SPEAKER_01: you know,
• 7:04
  SPEAKER_01: their veins are flat or they tense up and their arm shifts the exact second
• 7:07
  SPEAKER_01: the needle touches the skin.
• 7:08
  SPEAKER_01: What is the physical risk mitigation if the patient flinches or
• 7:11
  SPEAKER_01: the anatomy just doesn't cooperate?
• 7:13
  SPEAKER_00: The fail safes are layered incredibly deeply into the system.
• 7:17
  SPEAKER_00: First, the AI operates under strict physiological parameters.
• 7:21
  SPEAKER_00: So if the target vein doesn't meet exact mathematical criteria for depth
• 7:25
  SPEAKER_00: and diameter, the machine refuses to attempt the stick.
• 7:28
  SPEAKER_01: It just stops.
• 7:29
  SPEAKER_00: It just halts the procedure.
• 7:31
  SPEAKER_00: There is absolutely no scenario where the robot is fishing for a vein under
• 7:36
  SPEAKER_00: the skin like a frustrated human might do.
• 7:39
  SPEAKER_01: Oh, wow.
• 7:39
  SPEAKER_01: Okay, so it's very conservative.
• 7:40
  SPEAKER_00: Extremely.
• 7:42
  SPEAKER_00: Once the AI commits to a safe target,
• 7:44
  SPEAKER_00: a mechanical arm inserts a sterile 21-gauge needle at a precise angle,
• 7:49
  SPEAKER_00: between 10 and 30 degrees, executing with submillimeter precision.
• 7:53
  SPEAKER_00: And regarding patient movement,
• 7:55
  SPEAKER_00: the console is actually surrounded by high-frequency motion tracking sensors.
• 7:59
  SPEAKER_01: Okay, so if they twitch...
• 8:00
  SPEAKER_00: And beyond the visual and acoustic sensors,
• 8:02
  SPEAKER_00: there's tactile feedback built into the robotics too, right?
• 8:20
  SPEAKER_00: Yeah, the device utilizes onboard load cell sensors.
• 8:24
  SPEAKER_00: A load cell basically translates physical force into a measurable electrical output.
• 8:28
  SPEAKER_00: So as the needle traverses the epidermis,
• 8:31
  SPEAKER_00: the dermis, and subcutaneous tissue, The resistance profile changes.
• 8:35
  SPEAKER_01: Like it can feel how hard it's pushing.
• 8:37
  SPEAKER_00: Exactly.
• 8:38
  SPEAKER_00: The AI is trained on the normal microresistance curve of a human vein wall.
• 8:42
  SPEAKER_00: If the needle encounters tissue density
• 8:44
  SPEAKER_00: that falls outside those expected electrical parameters,
• 8:47
  SPEAKER_00: meaning it hits, say,
• 8:48
  SPEAKER_00: unexpected scar tissue or a dense anatomical anomaly,
• 8:52
  SPEAKER_00: it automatically pauses and alerts the clinical supervisor.
• 8:55
  SPEAKER_01: So the physical risk mitigation is really a trifecta.
• 8:57
  SPEAKER_01: You've got active acoustic monitoring,
• 8:59
  SPEAKER_01: passive physical load cell sensors and strict software constraints.
• 9:03
  SPEAKER_01: Right.
• 9:03
  SPEAKER_01: But, you know,
• 9:04
  SPEAKER_01: engineering theory,
• 9:05
  SPEAKER_01: no matter how robust, doesn't secure a de novo clearance on its own.
• 9:09
  SPEAKER_01: QARA and clinical professionals need to see hard patient data.
• 9:13
  SPEAKER_01: So how did this technology actually perform under clinical scrutiny?
• 9:17
  SPEAKER_00: Well, the foundational data comes from the landmark 8OPT clinical trial,
• 9:22
  SPEAKER_00: which stands for Autonomous Blood Drawing Occamization and Performance Testing.
• 9:26
  SPEAKER_00: It was recently published in the peer-reviewed journal Clinical Chemistry.
• 9:29
  SPEAKER_01: And this was a pretty huge deal, right?
• 9:32
  SPEAKER_00: Oh, massive.
• 9:32
  SPEAKER_00: It was a multi-center trial conducted
• 9:34
  SPEAKER_00: across leading clinical laboratories designed specifically
• 9:37
  SPEAKER_00: to test both analytical equivalence and operational efficacy in
• 9:41
  SPEAKER_00: a live outpatient setting.
• 9:44
  SPEAKER_01: So looking at cohort one,
• 9:45
  SPEAKER_01: which was 153 patients, the primary hurdle was proving analytical equivalence.
• 9:51
  SPEAKER_01: Because you can't just assume the robot handles the specimen exactly the same way a human does.
• 9:55
  SPEAKER_01: You have to prove the molecular integrity of the blood isn't altered by the machine itself.
• 9:59
  SPEAKER_00: Right.
• 10:00
  SPEAKER_00: So the researchers use passing Bablock regression analysis on paired samples.
• 10:04
  SPEAKER_00: And for those unfamiliar,
• 10:05
  SPEAKER_00: this is a highly stringent statistical model used in method comparison studies.
• 10:10
  SPEAKER_00: It essentially plots the results from the human-drawn sample against
• 10:13
  SPEAKER_00: the robot-drawn sample to ensure there is zero proportional bias between
• 10:17
  SPEAKER_00: the two methods.
• 10:18
  SPEAKER_01: And they tested some highly sensitive parameters.
• 10:20
  SPEAKER_01: I mean, they looked at coagulation markers like ABTT and PT and citrate tubes.
• 10:25
  SPEAKER_01: Because if the robot's needle creates excessive turbulence during the draw,
• 10:29
  SPEAKER_01: it can prematurely trigger the coagulation cascade,
• 10:32
  SPEAKER_01: which would completely skew those PT results.
• 10:35
  SPEAKER_00: Yeah,
• 10:35
  SPEAKER_00: and they also analyzed clinical chemistry markers like LDH and AST in serum tubes
• 10:41
  SPEAKER_00: and total platelet counts in EDTA tubes.
• 10:44
  SPEAKER_01: Right,
• 10:44
  SPEAKER_01: because if the robotic handling causes sheer stress that ruptures red blood cells,
• 10:48
  SPEAKER_01: those intracellular enzymes leak into the serum and your chemistry panel is ruined.
• 10:53
  SPEAKER_00: Exactly.
• 10:54
  SPEAKER_00: But across every single parameter tested,
• 10:56
  SPEAKER_00: the passing Bablock regression confirmed no statistically significant difference.
• 11:01
  SPEAKER_00: The clinical equivalence was absolute.
• 11:03
  SPEAKER_01: Which is incredible.
• 11:04
  SPEAKER_01: But that equivalence isn't just about how the needle enters the vein, right?
• 11:07
  SPEAKER_01: It's heavily dependent on how the machine handles the tube after it fills,
• 11:11
  SPEAKER_01: which is a massive pain point in the pre-analytical phase.
• 11:14
  SPEAKER_00: It really is.
• 11:15
  SPEAKER_00: Pre-analytical errors account for roughly 60 to 70 percent of all diagnostic errors
• 11:20
  SPEAKER_00: in the clinical lab.
• 11:21
  SPEAKER_00: And a huge portion of those errors stem from improper mixing.
• 11:24
  SPEAKER_00: I mean, manufacturer guidelines for tube inversions vary wildly.
• 11:28
  SPEAKER_00: Citrate tubes require 3 to 4 inversions.
• 11:31
  SPEAKER_00: EDTA tubes require 8 to 10.
• 11:33
  SPEAKER_01: And human phlebotomists are operating in these high-stress, high-volume environments.
• 11:37
  SPEAKER_01: They frequently just eyeball the process,
• 11:39
  SPEAKER_01: so they might shake the tube vigorously,
• 11:40
  SPEAKER_01: which creates sheer stress and destroys the red blood cells, causing hemolysis.
• 11:45
  SPEAKER_01: Or, you know,
• 11:46
  SPEAKER_01: they might forget to invert the tube entirely,
• 11:48
  SPEAKER_01: which leads to microclots that jam up those million-dollar analyzers we talked
• 11:51
  SPEAKER_01: about earlier.
• 11:52
  SPEAKER_00: Yeah.
• 11:53
  SPEAKER_00: But all that completely neutralizes that human variability.
• 11:56
  SPEAKER_00: The robotic system mechanically performs exactly nine standardized gentle inversions
• 12:02
  SPEAKER_00: for every single tube it draws immediately upon filling.
• 12:05
  SPEAKER_01: Wow.
• 12:06
  SPEAKER_00: It doesn't matter what color the tube top is or what the specific test is.
• 12:09
  SPEAKER_00: The robotic arm executes nine inversions at the exact same rotational speed every single time.
• 12:15
  SPEAKER_01: Think of it like CNC machining versus hand carving.
• 12:18
  SPEAKER_01: The CNC machine doesn't get tired or rushed or distracted.
• 12:21
  SPEAKER_01: It applies the exact same torque and angle on the thousandth part as it did on the first.
• 12:27
  SPEAKER_01: The robotic arm is applying that same industrial consistency to specimen handling.
• 12:31
  SPEAKER_00: And the trial data proved that this robotic consistency doesn't just meet clinical standards.
• 12:36
  SPEAKER_00: It actively standardizes a critical pre-analytical step
• 12:40
  SPEAKER_00: that is chronically mismanaged in manual workflows.
• 12:43
  SPEAKER_01: That leads directly into the efficacy data from Cohort 2 of the A.P.T.
• 12:47
  SPEAKER_01: trial, which evaluated the system across 1,633 patients in routine outpatient conditions.
• 12:53
  SPEAKER_00: So the primary efficacy endpoint was the first stick success rate.
• 12:57
  SPEAKER_00: When Aleta identified a suitable vein,
• 12:59
  SPEAKER_00: it successfully drew all required clinical tubes on the first stick 94.5% of
• 13:03
  SPEAKER_00: the time.
• 13:04
  SPEAKER_01: A 94.5% success rate is incredibly impressive.
• 13:08
  SPEAKER_01: I mean, that places it on par with or even exceeding highly trained human phlebotomists.
• 13:13
  SPEAKER_01: But the subgroup data is what genuinely demands attention from clinical directors.
• 13:16
  SPEAKER_00: Oh, absolutely.
• 13:17
  SPEAKER_01: In patients classified as obese with a BMI over 30,
• 13:21
  SPEAKER_01: the robot success rate actually jumped to 97.4%.
• 13:25
  SPEAKER_01: And in patients with self-reported difficult venous access or DVA,
• 13:29
  SPEAKER_01: you know,
• 13:29
  SPEAKER_01: the patients who historically endure three or four manual needle sticks,
• 13:34
  SPEAKER_01: The robot succeeded 92.7% of the time.
• 13:37
  SPEAKER_00: The machine is outperforming human phlebotomists in notoriously tricky demographics
• 13:42
  SPEAKER_00: because it operates entirely outside the limitations of human senses.
• 13:47
  SPEAKER_00: A human phlebotomist relies on sight and tactile palpation.
• 13:51
  SPEAKER_00: But in an obese patient, the vascular structure is often buried beneath adipose tissue.
• 13:56
  SPEAKER_01: Right, you can't see it.
• 13:57
  SPEAKER_00: The human cannot see it and often cannot feel it.
• 13:59
  SPEAKER_00: They're essentially relying on anatomical landmarks and making a highly educated guess.
• 14:03
  SPEAKER_01: But the robot sensors don't care about adipose tissue.
• 14:06
  SPEAKER_01: It's literally like having a submarine sonar system mapped onto the needle.
• 14:09
  SPEAKER_00: That's a perfect way to describe it.
• 14:11
  SPEAKER_00: The 3D ultrasound waves penetrate the lipid layers seamlessly.
• 14:14
  SPEAKER_00: Sound waves reflect off the fluid dynamics of the blood and the dense wall of
• 14:18
  SPEAKER_00: the vein,
• 14:18
  SPEAKER_00: presenting a crystal clear acoustic boundary on the machine's imaging matrix.
• 14:22
  SPEAKER_01: So it's not guessing at all.
• 14:24
  SPEAKER_00: Not at all.
• 14:25
  SPEAKER_00: For the AI,
• 14:25
  SPEAKER_00: a deep vein in an obese patient is mapped with the exact same mathematical precision
• 14:30
  SPEAKER_00: as a superficial vein in a thin patient.
• 14:33
  SPEAKER_00: The guesswork is entirely removed.
• 14:35
  SPEAKER_01: And when you remove the guesswork, the safety profile naturally improves.
• 14:38
  SPEAKER_01: Across those 1,633 patients,
• 14:41
  SPEAKER_01: the adverse event rate was just 0.6%,
• 14:44
  SPEAKER_01: and every single event was classified as mild,
• 14:46
  SPEAKER_01: such as a minor hematoma or vasovagal syncope,
• 14:49
  SPEAKER_01: which is simply a patient fainting from anxiety.
• 14:52
  SPEAKER_01: Which, let's be honest, is a physiological response that occurs in manual draws daily anyway.
• 14:57
  SPEAKER_00: Right.
• 14:57
  SPEAKER_00: And the hemolysis rate,
• 14:58
  SPEAKER_00: which is a critical quality metric for any lab director, was recorded at just 0.3%.
• 15:03
  SPEAKER_00: This is exceptionally low.
• 15:05
  SPEAKER_00: In fact, during the trial, there was a Oh, right.
• 15:13
  SPEAKER_01: This was fascinating.
• 15:15
  SPEAKER_01: So they conducted an ad hoc review of the manual phlebotomy samples being drawn
• 15:19
  SPEAKER_01: by humans at that exact same site during the same time period.
• 15:22
  SPEAKER_00: And the manual human-drawn samples at that facility had a hemolysis rate of 6.9%.
• 15:28
  SPEAKER_00: So the slight elevation wasn't a mechanical failure of the robot.
• 15:32
  SPEAKER_00: The robot actually exposed a localized human error in sample handling practices at
• 15:38
  SPEAKER_00: that specific clinic.
• 15:39
  SPEAKER_01: Wow.
• 15:40
  SPEAKER_00: Yeah,
• 15:40
  SPEAKER_00: the machine's perfect consistency established a baseline that highlighted the flaws
• 15:45
  SPEAKER_00: in the manual workflow.
• 15:46
  SPEAKER_01: Okay, so the clinical data verifies the engineering,
• 15:49
  SPEAKER_01: and the FDA clearance verifies the regulatory framework.
• 15:52
  SPEAKER_01: But integrating a disruptive medical device
• 15:55
  SPEAKER_01: into massive health systems requires navigating the complexities
• 15:58
  SPEAKER_01: of human psychology.
• 16:00
  SPEAKER_01: I mean, you can engineer the perfect robotic arm,
• 16:03
  SPEAKER_01: but if a patient refuses to let a machine insert a needle into their vein,
• 16:07
  SPEAKER_01: the commercial momentum stops immediately.
• 16:09
  SPEAKER_00: Completely true.
• 16:10
  SPEAKER_00: And the engineers at Vitestro anticipated that psychological friction.
• 16:13
  SPEAKER_00: They designed the Oleta kiosk so that the mechanical tube switching assembly and
• 16:17
  SPEAKER_00: the needle itself are completely hidden from the patient's line of sight.
• 16:20
  SPEAKER_01: It leverages the whole out of sight, out of mind principle.
• 16:24
  SPEAKER_01: The patient never sees the sharp edge of the needle approaching their skin,
• 16:27
  SPEAKER_01: and they don't sit there watching multiple tubes rapidly filling
• 16:31
  SPEAKER_01: with their own blood.
• 16:32
  SPEAKER_00: And the data strongly supports this psychological shielding.
• 16:35
  SPEAKER_00: In the NLPT trial,
• 16:37
  SPEAKER_00: 90% of patients reported that the robotic draw caused less pain,
• 16:42
  SPEAKER_00: or an identical level of pain,
• 16:45
  SPEAKER_00: compared to their historical experiences with human draws.
• 16:47
  SPEAKER_01: That's huge.
• 16:49
  SPEAKER_00: It is.
• 16:49
  SPEAKER_00: And when asked about future use,
• 16:51
  SPEAKER_00: 82% either actively preferred the robot or had no preference between the robot and
• 16:55
  SPEAKER_00: a human.
• 16:56
  SPEAKER_01: And Vitestra also ran a simulated use study in the United States at the Mayo Clinic,
• 17:01
  SPEAKER_01: where 86% of patients indicated a willingness to use the Aleta device.
• 17:05
  SPEAKER_01: So that high level of patient acceptance is really the catalyst for
• 17:08
  SPEAKER_01: the commercial integration we are seeing right now.
• 17:11
  SPEAKER_00: The industry buy-in is very substantial at this point.
• 17:14
  SPEAKER_00: Matestro recently closed a $70 million Series B funding round,
• 17:18
  SPEAKER_00: and the cap table includes strategic investments from the LabCorp Venture Fund,
• 17:22
  SPEAKER_00: the Mayo Clinic, Sutter Health, and notably, Dr. Fred Maul.
• 17:26
  SPEAKER_01: Right.
• 17:26
  SPEAKER_01: And Dr. Fred Maul is the pioneer of robotic surgery.
• 17:29
  SPEAKER_01: He co-founded Intuitive Surgical, the company behind the DaVinci robots.
• 17:34
  SPEAKER_00: Exactly.
• 17:34
  SPEAKER_00: When a pioneer of surgical robotics backs an autonomous platform like this,
• 17:38
  SPEAKER_00: it signals a major shift.
• 17:40
  SPEAKER_00: He's drawn direct parallels between the early skepticism surrounding robotic surgery
• 17:44
  SPEAKER_00: and the current state of autonomous phlebotomy,
• 17:47
  SPEAKER_00: really highlighting its potential to redefine
• 17:49
  SPEAKER_00: an entirely under-innovated clinical workflow.
• 17:52
  SPEAKER_01: And they are moving aggressively to validate this in the U.S.
• 17:55
  SPEAKER_01: market, partnering with Northwestern Medicine.
• 17:57
  SPEAKER_01: Dr. Gregory Retzinger and Dr.
• 17:59
  SPEAKER_01: Gary Noskin are actively working to validate how that one
• 18:02
  SPEAKER_01: to three supervised ratio holds up under the chaotic high volume realities of
• 18:07
  SPEAKER_01: an American outpatient center.
• 18:08
  SPEAKER_00: Yeah.
• 18:08
  SPEAKER_00: And Dr. Tyus Van Holten,
• 18:10
  SPEAKER_00: who is one of the clinical chemists involved in the European trials,
• 18:13
  SPEAKER_00: provided a highly relevant framework for this whole transition.
• 18:17
  SPEAKER_00: He compared the integration of autonomous phlebotomy to the evolution of
• 18:20
  SPEAKER_00: the supermarket self-checkout.
• 18:22
  SPEAKER_01: I love that analogy,
• 18:23
  SPEAKER_01: because when self-checkout kiosks first appeared,
• 18:26
  SPEAKER_01: consumer skepticism was incredibly high.
• 18:29
  SPEAKER_01: Today, they handle the vast majority of standard transactions.
• 18:32
  SPEAKER_01: But the human cashiers were not eliminated.
• 18:35
  SPEAKER_01: The automated kiosks absorb the high volume standard workflow,
• 18:39
  SPEAKER_01: while the humans manage the complex edge cases and provide vital oversight.
• 18:43
  SPEAKER_00: Exactly.
• 18:44
  SPEAKER_00: Oletta serves as the high-throughput kiosk for routine diagnostics.
• 18:48
  SPEAKER_00: It allows highly trained human phlebotomists to step away from the relentless cue
• 18:52
  SPEAKER_00: of standard outpatient draws.
• 18:54
  SPEAKER_00: They can redirect their expertise to pediatric cases,
• 18:57
  SPEAKER_00: extreme trauma patients in the ED,
• 18:59
  SPEAKER_00: or just individuals who inherently require a human touch while
• 19:02
  SPEAKER_00: the robot ensures flawless pre-analytical consistency for the masses.
• 19:06
  SPEAKER_01: So wrapping up,
• 19:07
  SPEAKER_01: the FDA's de novo authorization of the Oletta ARPD represents
• 19:10
  SPEAKER_01: a massive operational paradigm shift.
• 19:12
  SPEAKER_01: By merging multimodal imaging with rigid safety engineering,
• 19:16
  SPEAKER_01: this technology is establishing a totally new standard of care.
• 19:19
  SPEAKER_01: It mitigates physical risks,
• 19:21
  SPEAKER_01: standardizes the chronically mismanaged pre-analytical phase,
• 19:24
  SPEAKER_01: and offers a highly scalable solution to a global labor shortage.
• 19:28
  SPEAKER_00: And, you know,
• 19:28
  SPEAKER_00: the immediate focus is naturally on high throughput centralized hospital labs,
• 19:34
  SPEAKER_00: as that is where the economic and operational pain points are most acute right now.
• 19:39
  SPEAKER_00: However,
• 19:39
  SPEAKER_00: considering the 86 to 90 percent patient acceptance rates,
• 19:43
  SPEAKER_00: the next logical frontier for this technology is total decentralization.
• 19:48
  SPEAKER_01: Oh, that's an interesting thought,
• 19:49
  SPEAKER_01: because if the machine autonomously handles the complex imaging,
• 19:53
  SPEAKER_01: the rigid safety protocols,
• 19:55
  SPEAKER_01: and the sample standardization,
• 19:57
  SPEAKER_01: you no longer need a dedicated phlebotomy staff at every single point of care.
• 20:01
  SPEAKER_01: Imagine a near future where these autonomous kiosks are deployed
• 20:04
  SPEAKER_01: in local pharmacies,
• 20:05
  SPEAKER_01: neighborhood general practitioner offices, or even rural health clinics.
• 20:09
  SPEAKER_01: It fundamentally shifts where, when, and how humanity accesses diagnostic medicine.
• 20:14
  SPEAKER_00: It really does.
• 20:14
  SPEAKER_00: It takes the lab to the patient in a way we've never been able to before.
• 20:17
  SPEAKER_01: Well, for everyone listening,
• 20:18
  SPEAKER_01: we hope this deep dive provided the rigorous clinical
• 20:21
  SPEAKER_01: and operational insights you need to navigate this rapidly evolving landscape.
• 20:25
  SPEAKER_01: Thank you for joining us as we explored the future of pre-analytical automation.
• 20:29
  SPEAKER_01: Until next time.