AI is Now Inside You: Ingestible Robots & The Future of VA Systems

With AI again and with micro robotics pills can be ingested and travel around the the body to go and tackle a particular area that is sounds very futuristic but there are companies working on that

how you're using AI in your daily life

so that's one area where already AI is saving lives potentially by looking at a mamogram and knowing this is requires immediate action or or not AI is um starting starting to help not in the core of the project but again at the beginning

and at the end.

AI is it hype or help? Good day and welcome to the show. I'm your host Pete Cooper and our topic is AI hype or help. Today's guest is Miguel Ado.

Miguel is the president and CEO of Valair Systems, a tech development company operating out of Silicon Valley. And they have been in business for, would you believe, 47 years. So, Miguel has some, you know, really interesting insights into technology, particularly around consumer electronics and medical devices, and I'm really looking forward to hearing what he has to say. So, let's get into it. Hey, Miguel. Thank you. Welcome. Welcome to our podcast. Really nice to see you again.

Likewise. It's good to see you, Pete.

Yeah. And so, we're going to talk a little bit about your background and history and CES and, um, I think there's this is going to be really interesting.

So, Miguel um, from Volair systems, do you want to just introduce yourself, what you do, what your role is over at Valair?

Yeah, very well. I, um, my name is Miguel Dao. I'm the president CEO of a company called Valer Systems. We do electronic design and development for medical device companies, for wearables, for IoT companies. I've been here four years. Uh the company has been around for 46. It was founded in 1979 uh in Sunnyville, right in the epicenter of uh

Silicon Valley. Uh the founder still uh talk to him three times a day, Walt Mcclelay. You've met him, Pete.

Um, and we're delighted that this month, earlier this month of February, uh, 2026, we sold the company after 46 years as an independent company to, uh, an organization out of Montreal, Canada called Innovabot.

Um, and this is their first acquisition in US. They invest and acquire in Canada. But, uh, I'll be delighted. We're all staying on board, Walt and myself and our teams. Yeah.

And, uh, yeah, great, uh, energy all around. So what does Volair systems do?

Hardware and firmware primarily we focus on medical devices classes one two three in the FDA scale we do a lot of wearables uh both for medical and non-medical purposes wellness and industrial quite a bit of IoT internet of things a little bit of consumer electronics but not as much we won't do mechanical we won't do industrial design uh we have partners for that but our engineers we have a deep bench of well seasoned engineers that can do anything under the sun with sensors, motion control, power management, wireless connectivity, analog digital combinations, and more now we're starting to look at physical AI uh which will be I'm sure will come up in this conversation.

So would there be a product out there that um we could focus the conversation around that you could talk about incorporates all those things? I mean after all those years there's probably many many products. What's one that I guess the audience might even know of?

Yeah. Yeah. And it's interesting. I'd have to go back to the NVAs and see what I'd a specific product that I'd be able to share because very often, Pete, it's interesting. We are added to an engineering team as a staff augmentation or a skill uh subject matter experts. But very often in our NDAs, we're not allowed to communicate. But what I can say in generic terms is cardio and neuro are very interesting spaces for us.

We're very unique in that space in terms of managing cardiovascular conditions. Managing neurological conditions, Parkinson's or depression or other mental conditions, sleep apnea, spinal surgery. We do a lot in robotic surgery as well. So, a lot of really invasive and also not invasive, anything that you can wear on your finger, on your wrist, on your clothes, smart textiles. We do quite a bit even with smart hearables and we do a lot with uh um also um yeah I mentioned robotics. Robotics is a space that's really picking up but mostly these would not be consumer products as such. We're not doing aura rings. We're not doing uh I Apple watches cannot compete with those.

We're typically helping either startups or more established companies mostly in the uh medte space and very often it is the development of a prototype or the device itself. We don't manufacture anything but we help them in the design and then they run with that ball.

So would a pacemaker fall under that uh

yeah category?

Anything that has a a battery that has sensors that has some sort of uh PC board or or chips, etc. That that's the world that we touch. Yeah.

So, so if anyone doesn't know what a pacemaker is, um it's something that's embedded.

It's a piece that's embedded around the heart and then there's usually a piece on the outside, right, which uh which is battery powered and it communicates between the two of them and it helps to regulate the pace of the heart because I think some people have this irregular heart rate heartbeats and that helps to regulate that. Yeah.

All right. I think that might be something that mo that most of us could relate to, but um you mentioned FDA classes 1, two, and three. Um I think that's got something to do with whether it's embedded under the skin or whether it's on the surface. Is that right?

That's right.

It's basically having to do with the level of of risk and the the level of of seriousness if it goes wrong. And this is something that will come up in our AI conversation as well. It's okay for AI to hallucinate in in a simple analysis or a Shakespearean sonnet. It's not okay if it elucidates when you have a pacemaker inside or as we work in automotive and green energies and um we have devices in the space station. You cannot have hallucinations there and you have to be careful. But to your point, yes, class one is something non-invasive. Class two becomes more serious. Class 3 could be literally life or death.

Wow. Oh, okay.

So,

yeah,

it's also the not just the fact that it's embedded under the skin, it's how critical it is to someone's someone's health and life and so on. Okay. Wow. That sounds uh like a tough challenge. Um even at a just basic engineering level because it's it's like very deterministic. You got to have it exactly right. It's got to be reliable. I mean, is there uh is that a challenge for the AI or is that just completely out of the park or you seeing it coming in?

Um, yeah. So, we were joking about it offline when you invited me to your podcast. I'm thrilled to be here. But I'm one of those folks that h has been particularly conservative.

I I'm a generally a lagard in terms of technologies. I remember the do for those viewers who are old enough to remember the dotcom boom and bust, right? There was peepod and webvan and pets. Com and boo. Com all these

big companies and they're all gone right and so it's very interesting we'll see the same similar shakeout here now but we're right in the middle of the storm and so it's hard to differentiate who will you're already starting to see some winners you're starting to see some some some movement there but um AI in general um because we work in the physical world of electronics of devices of hardware and electronics. Yeah.

Um it's not there just yet.

And so it will take a while, but it's coming. It's coming. We're seeing the uh um the beginnings, the seeds of uh something uh quite exciting slash scary if not managed properly.

So our audience might be familiar with is is some of the trials and tribulations that Tesla's been through. You know, there's been a promise from Tesla for many years that we're going to have um self-driving cars and and they doubled down on AI vision systems. And then there was some really bad accidents, right, where the vision system just didn't see uh a truck, for example, and the car went under the truck.

It thought it was blue sky.

And it doesn't seem to be that the the AI vision systems have really got there yet to the point where they're reliable. And I think the other thing that I mean I know we're talking about the auto industry but I see I want to see if this relates to the medical side of things is when someone drives a car they're responsible right and if they make a mistake then you know there's fines and accountability but with AI who's accountable right

that's right yeah and so that's a great example on the automotive side with with with Tesla and other vehicles and other examples that we've all seen and read about in the medical space in the medical device space that's only amplified, right?

You got you have to have the human in the loop otherwise it's just you cannot trust it between

I mean we read these things about not just hallucinations but outright lies AI is actually learning to actually evade um the human intervention.

So we have to be very cautious because there's AI which is more kind of what we've seen with ChatGPT kind of responding to prompts and now there's agentic AI right where it's actually

bypassing the human intervention alto together and just proactively leading what we've seen in our world is that this notion of vibe coding Pete where you just kind of create something on the fly leveraging the AI tools as as as sophisticated and as much as it has improved. We cannot use it for medical devices or indeed for many of our wellness and industrial devices because what we find is when we get AI initiated um schematics and block diagrams and MVPs minimum uh vio v viable products etc.

We have to start from scratch. Right now I'm saying this is changing rapidly but up until now we have not been able to take something that has some sort of AI intervention other than very minimal research done or maybe some work on the product requirements uh doc document the PRD some of this more

I don't want to belittle it and call it superficial but certainly not at the device at the hardware and firmware AI cannot write firmware firmware is defined as embedded software software that's inside a a pacemaker to make it do things.

And so um we're yeah it's it's going to be a journey for sure and we cannot afford the AI does not yet have the I'm not going to call it intelligence but the uh the the the critical thinking and judgment and ability to uh to differentiate risky situations from uh acceptable situations.

Right.

So it sounds like um you could use it in your industry for maybe doing some research into uh you know what's available what might be some common approaches to a particular technical problem helping with some of this the I guess the planning side of things writing requirements but then at some point you've got to say well we got to actually build this thing and that means we got to start from engineering principles the the old fashioned way, right? Where you get someone who knows what they're doing and builds the thing from the ground up, right?

I think that's a good way to put it.

The way I think about it now, simplistically, oversimplistically,

is that AI can help at the beginning and it can help at the end,

but it cannot right now help, and I'm going to emphasize these words right now because it's changing rapidly. As we speak, things are being launched and improved and it's the speed of change is overwhelming. But in the middle today,

you need the human and um so certainly at the very beginning basically the LLMs and the AI tools can help um automate and accelerate and identify opportunities early on and minimize that laborious and arduous and timeconuming uh process.

At the end it can help certainly prepare your documentation for your FDA submission for instance compliance and regulatory and all that because it just follows everything that's been done uh since beginning of time. It knows

all the submissions to the FDA since the FDA started and it can help us. It can help the client. Can help us as the uh the partners uh collaborators in in providing a more accurate submission that perhaps has fewer chances of being rejected or um

pushed back for whatever reason. But in the middle, the actual construction and building [clears throat]

of the cathedral, it needs to be still very much by humans.

I wouldn't have you said that it could be more accurate. That's interesting, isn't it? That um I would have thought you I would thought we were going to go down the path of it's going to be faster and which which means you need less people to create the project documents and the the starting pieces and and also to be faster at the end. But you actually brought in a an idea I hadn't thought of which is actually it's more accurate because it can look at many many prior submissions whereas a human can do that they can maybe have experience

right

but they still could miss things.

They're not looking at it actively right there and then to pull this you know um the I guess the if you look at lots of successful ones and you train a model to look at the successful ones then it's going to follow that path just the same way a human would but it takes so many years and so much time and energy to do that

it's very hard to get a human to read you know 20 or 30 submissions and then come and write one

right exactly and so to be clear. Yeah. When I speak of accuracy and I speak of uh um the ability to deliver a finished product that's acceptable say to the FDA or to whatever organization.

Yeah. Right now I'm referring to the written documents.

I'm not I still do not believe that AI can uh can do the

accurate uh bill of materials, the accurate construction of the PC board, the act accurate design of the um of the device or the prototype, but certainly in the uh in the product requirements part, the market requirements or the uh the tail end of that that whole cycle as well. They can be very helpful.

Yeah. And and we'll see how this evolves, but it needs some serious guard rails.

Yeah. So, I mean, from engineering perspective, uh we're talking about that traditional approach, and it's starting to be challenged in some ways that I became aware of recently where um let's take a a motorcycle.

They wanted to build the the tension arm, the torsion arm, which is holds the rear wheel in place. And they said to this AI generative tool, create lots of different types, right? And then analyze the stress of each of them and it would go and do hundreds of different types of variants. And this is some weird shapes, beautiful designs that just that look crazy. It's not just like a torsion bar. It's like intricate design.

Analyze that and then do and do hundreds of them and then then pick the best ones and then analyze that one and then build it and build it using 3D additive techniques. Has almost taken the human out to the point of building that swing arm for

Yeah.

for the bike.

Almost taken them out of that. [snorts]

Is there anything like that happening in your space?

Not not because we're primarily medical devices and very sophisticated and highly regulated non-medical spaces um oil and gas, automotive, aviation, aerospace, green tech, eggtech, agricultural technologies. M

um certainly I can see for a part in a uh in a a bike or perhaps a non uh dangerous non-lethal component in some sort of uh equipment

or or machinery. It it probably makes a lot of sense with also

the advents in robotics in 3D printing and all that. Uh in our world honestly it's we're starting to get exposure to it.

And don't get me wrong, on the marketing side, on the business development side, we are

AI first right now in in identifying lookalikes for our lead generation and creating some AI generated or AI supported content. Still needs to look human. Can't be AI slop.

We see that in the commercials during the Super Bowl, right? So, you got to be careful you don't go too far because people are always are also there's a backlash. It's very interesting. Not just with my kids generation. I have two kids in high school who are completely AI already even though they use some of it for school work. It's very interesting.

They will not I'll show them a cute little cat video

and they'll just tell me that's not interesting. That's not real. It's not genuine. And by the way, Dad, they're taking away uh work and business from creatives and videographers and and musicians. So, it's a very interesting that will be a different podcast. But this generation has a much um um more mature attitude about this that perhaps even we do. But

certainly um on the marketing and business side, we're finding it very useful on the actual hardware, firmware, electronics development and design. It needs to be still very much humanled at the core of the the project. Yeah.

Yeah.

That's uh that's kind of refreshing to to hear that because I think what's happened in the last few years with unless you've been under a rock you you're using the the LLM's um you know ChatGPT of course and Claude and Gemini those those tools everyone's familiar with that's the generative side and there's this element that's what AI is um but of course AI has been around for a long time but in a different way and that is more on the deterministic side of things which is sort of where my background was in the AI vision systems where you you want to resolve uh what the camera sees and it generates classes right so it's deterministic it's saying okay I can see lots of things and those are

things that I can see in those locations that's to to me there's almost there's two worlds of AI generative making new stuff creating new stuff having conversations and chat bots and things so where there has to be a creation part but then there's the deterministic where you're kind of resolving down to something I think that's where we might see some um some of AI coming into your world, right? Where for example, if the heart rate is is so erratic um it could analyze that with AI and and suggest a a remedy.

Yeah, I agree. It's it's interesting your your um words just now just pro prompted some thoughts. It's interesting because Stephen Spielberg did a movie called AI back in the 80s.

I mean, this notion has been around forever, right? Yeah,

of course we've had all the dystopian movies starting with Kubri back in the uh what was it late60s I believe 2001 Space Odyssey an amazing visionary movie but but it's very interesting I mean you and I remember IBM Watson right uh that was like one of the big finally corporate pushes into the precursor of AI so it's been around for decades and decades

and it has been

um something that I since I am based well I actually live in San Diego but our organization is based in in Silicon Valley.

Up there every other week and I see the energy that's happening both in the fundraising side certainly but also in the actual um corporate and innovative side with all these startups and scaleups popping up here in the the future of AI will actually be uh a really interesting mix of the Google's and uh uh SpaceX's or XAI however it's called now of the world but also these upand cominging organizations many of which have not even started and it's um I I find the moment to be we're at a crossroads. We are um at a moment where yes, I am finally starting to see the potential for AI and you talk about deterministic versus reactive.

The the AI that we've all kind of grown accustomed to in the last four years was more write a prompt and it'll come back with something. But now we will start seeing indeed AI taking that deterministic or that agentic or that proactive approach. It'll take still a little bit longer to uh materialize itself in the physical space in the uh uh let me say in the high-risk uh life or death um regulated and um hyper sophisticated space and medical device is one space. There are many others but that's the one I touched directly. Um yeah we'll see where we go from here. I'm ambivalent still about I'm optimistic.

I do believe that

many cancers for instance will be resolved and go away in the next decade maybe even sooner. I do believe that we're seeing that's one concrete area where I see value right now Pete is in

AI is able to look at a X-ray at an MRI at a some sort of imaging uh report and determine if a cell is cancerous or benign better than humans better than

radiologists and surgeons with 30 years of experience. That's astounding to me. We're seeing those facts. So that's one area where already AI is saving lives potentially by looking at a mamogram and knowing this is requires uh immediate action or or not.

And so

uh we have these budding uh shoots of of uh of excellence coming out but it's still very much a mixed uh uh scorecard.

It is and I think that it's a good example of the the imaging. And I think what you you still have your radiologist, you still have your GP, you still have your specialists. Obviously, you still have this the patient. So there's still the need for this core of human pe humans that that are going to be needed. But now maybe we just amplify the capability through AI. And I think this is to me this is the revolution that's coming. It's how people ultimately because companies like yours and mine are organizations of people. That's what that's what a company is.

Um it's we're still people and we but how can we use AI? And I think it's it's like in that instance it's kind of like a really good example of a tool, right? You you you take the existing organization of those five people, the radiologist, and they add that AI tool in to supplement their work. And it can even help them get be better at their job because they the AI sees things they don't and then they learn from it. [sighs] Um, and I'm really a big fan of that. I think that the amplification of our capabilities through AI, I think is the next big step. I think it's the the transformation that every business is going to witness and hopefully embrace for their betterment.

I agree.

I I I do. It's interesting. You mentioned CES at the beginning of the the chat and um I was at CES this year. I believe you were there as well.

I didn't go this year. No, talk about what you saw there actually.

Well, what I saw is AI everywhere. AI. AI. AI everywhere. And I saw it already a little bit last year. CES is a big It used to be called Consumer Electronic Show. It's now consumer and industrial and medical and it's all uh things electronics. And it's a a zoo. It's 130, 000 people descend in Las Vegas and the the convention center and the Venetian expo or uh wallto-wall AI everywhere. It's now become as ubiquitous. It's like us saying now here, Pete, my company has a website.

Of course, it does. My laptop as an onoff button. It is. It's just It's par for the course. It's a given. Everything. What I saw this year was interesting. I attended the um the keynote. I believe it was at the Spear. I attended different events at the Spear. Was this one at the with Dell, the uh SVP of Dell computers, my former competitor. I worked for Hila Packer for 12 years. He did not mention AI or in fact what he did say was if I recall um we've learned that people actually don't want too much AI in their laptop. So we're we're scaling back.

You know they they're accustomed to a certain u out of box experience or a certain kind of user interface UI experience and they've actually scaled back a little bit. Going back to my point about my children kind of overreacting or not overreacting but um um um backing away a little bit from the AI onslaught. So it's it's very interesting both you mentioned people. It's

all starts and ends with with humans. We

we're this is not two AI robot overlords having a podcast. Might be interesting profile, but

that's right. Who would care? Who the kids

they they they don't sweat. I'm sweating a little bit because I'm just in the moment, right?

You can tell that we're connecting on a visceral level and that cannot be replicated. So, it's very interesting what's happening and we just need to navigate this brave new world um with caution with uh a healthy dose of uh of trepidation but also with excitement because if we can save lives, if we can save the planet, not if but as we save lives, make lives healthier, people healthier and live longer and help the planet and and help

uh equality around the world.

I know it sounds like a little kumbaya but it's there's I see the power of AI and I certainly see it in my world as well of medical devices already and there are several other projects I can even talk about uh under NDA as well but we are talking about Parkinson's and stroke and um um perhaps managing um brain cancers in a way that half of the brain does not need to be excised to and and unfortunately remove healthy cells and healthy parts of the body to get to something. So AI is um starting starting to help not in the core of the project but again at the beginning

and at the end. That's how I

was there anything that blew your mind at CES.

Uh [gasps] I joke with somebody that CES is perfect for tech bros. A lot of uh uh interesting uh drone technology and automotive technology, a lot of smart home stuff. I don't think we're going to have a robot at home doing our laundry or ironing our clothes just yet,

but I am certainly very excited about the nanotechnologies that are coming um where uh

with AI again and with micro robotics pills can be ingested and travel around the the body to go and tackle a particular area. That is sounds very futuristic, but there are companies working on that.

Um, on the wearable side as well, of course, we wear our rings to measure the quality of our sleep and to measure our uh blood rate and and uh oxygen, but there'll be wearables that can actually predict. You're going to have a a medical episode or um we have to be careful as well because you don't want to be told by a wearable that you might have propensity for something. You want to uh really have that conversation with your doctor. But there's a lot of um um medical advents in just around the corner

that again will uh really make a difference in our lives.

Yeah.

Well,

we're going to sort of start sum summing up, but before we do that, I'd like to know sort of how you're using AI in your daily life. I mean, are you using it for work? Are you using it at home? Man,

I've resisted it for the longest time. Like I said, I'm very much a lagard. I I used to be leading edge, Pete, to the point where I became bleeding edge

and and and there's a fine line um with that. And I've kind of learned my lessons whether it's uh new technologies or uh buying a new uh shiny object or um I even have ambivalent opinions about cryptocurrencies and bitcoin. We can have another podcast on that.

But uh and so in my personal life, very rarely other than to write a beautiful uh poem for my wife or something like that at Christmas time. Um

nice. Yeah.

Um my kids are using it to help with their homework. My wife certainly uses it for her her job and I see her

becoming a big advocate of Claude used to be a ChatGPT. She writes books and it helps with the editing of the books and the uh the selection of of of titles and covers and and and so forth. So, we're starting to see it in my family. Ironically, I'm the engineer [clears throat] and I'm I'm the head of the uh the the the kind of uh technology and innovation company here, but I'm following the lead of my my family.

It kind of makes sense. I think in a way you

you you there's an almost an expectation that you you do resist it because

you have to stand behind something which is has to be medically safe and you have to have make sure it's proven and in a way I can understand your position very much so I mean it's not like you're an advertising you heading up an advertising company where you know if you're not using AI you're not in the game You have you have to play a different game. You have to be much more conservative.

I mean, I use the airline industry when I think about safety because I worked in that industry for a while.

Nobody wants to get on an airplane where anybody has taken any risks, right? Not even not not the pilots, the captain, the nobody. Everybody wants to be 100% sure that every safety precaution is made and let's not throw AI in into see if the engines will work better with AI or have an AI pilot up the front. God's sake, you know.

That's right. That's actually a great analogy. Yeah, I think of designing and developing medical devices much like uh the airplane experience you mentioned. You can certainly use AI to find the best flight uh for the best day at the best cost. You can certainly use AI to then um select the your favorite movie for your long flight from the the console.

You can certainly use AI um to track your luggage and to go through customs when you land at your destination, but in the flight with the pilots and the cockpits and the engines, you don't want to mess with that. AI is probably I'm sure not probably, I'm sure it's working. Uh Airbus and Boeing are using AI to make planes uh safer and faster and more environmental environmentally friendly and to use less fuel. I'm sure that's all happening for future Dreamliners and for future uh aircraft, but in the moment in the air, you do not want to tinker and find out that AI needs to reboot or AI started hallucinating and decided that is not really.

I thought the runway was there but it was a cloud or vice versa.

Exactly.

Yeah. Well, thank you Miguel. This has been great. So, final question for you. AI is it hype or help?

Yeah, I I think my uh my sentiment throughout this half hour with you has been pretty consistent. Um I believe it's been hype until now.

Uh other than some corner cases in other industries, etc. In in my space, medical devices primarily uh heavy sophisticated industrial spaces mostly hype. Um

I can see it not yet but I can see it quite soon becoming a help and already it's helping on the edges at the very beginning of the process at the very end of the process on the edges of an initiative.

I still believe we need those engineers with 35 years of experience to build that device and those um boards and and and those schematics. But um mostly hype until now

um it'll start becoming helpful.

But on the other hand, if governments and uh the the the folks in charge in the private and public sector are not cautious,

the help could backfire and we could end up in a quite a uh problematic scenario in the future. So, we need to be careful as a society.

I think so, too. Well, on that note, thank you, Miguel. Thank you very much for coming and chatting with us and uh I've really enjoyed it.

Yeah, I did too. Thank you so much, Pete. Thank you.

Bye.

Well, that was a very interesting discussion we had with Miguel where we dug into an area that I think most of us probably wouldn't be aware of um and how deep it really goes in terms of engineering. So, is AI relevant in that world? In the world of medical devices? Well, actually there is, but it's not where you might think. The core, the engineering core is still very much people with experience building uh products and building them with engineering methodologies that have been around for a long time. But at the front end of projects, we're um they're working out what to build and how to build and build up the the the start of the project.

And at the end of the project where they're creating the the submissions um to the FDA regulators then those two areas it seems that AI is becoming helpful. Well, if you want to know more about VA systems follow the links in the description. If you want to know more about Skillion please also do the same. I hope that helps and thanks for watching.