Episode 24
The Swiss Way - Liliane Paradise with Miromico
Liliane Paradise, COO of Miromico, talks about building industrial-grade LoRaWAN devices with Swiss precision, pioneering battery-free sensor systems, and the business case for hybrid battery configurations in critical environments. She shares how Miromico approaches product development from concept to mass production, prioritizing modularity, testability, and long-term sustainability.
- Explains why gateways are the most critical—and often overlooked—component in a LoRaWAN networ
- Breaks down the design philosophy behind Miromico’s modular gateway platform that supports LoRaWAN, BLE, Wi-Fi, and Mioty
- Highlights the business drivers for deploying energy harvesting sensors in schools, hospitals, and office buildings
- Details how Miromico launched the Mirror Inside Lux sensor, a fully battery-free device powered by indoor light with a 60-day supercapacitor backup
- Discusses the limitations of energy harvesting in critical use cases and why hybrid battery systems are the practical path forward
- Walks through Miromico’s product development lifecycle, from fast POC builds to customer-validated MVPs and scalable manufacturing
- Shares how design for testability and manufacturability saves cost during production—every screw, every second matters
- Explores the role of interoperability and standardization in reducing IoT deployment costs
- Explains how the European Cyber Resilience Act will impact IoT device security and reshape manufacturing timelines
- Offers insights into the complementary roles of LoRaWAN and Mioty in building future LPWAN solutions
Links:
Miromico website
- Helium Global IoT Coverage - Want to know if Helium coverage exists where you need it? Check out this map!
- Helium Foundation - The Helium Foundation's IoT Working Group (IOTWG) has generously provided support for the first 6 months of shows, please go check them out and consider using the Helium LoRaWAN as a primary or backup on your next deployment. With over a quarter million gateways deployed worldwide, it's likely that you have and can use Helium coverage.
- Support The Show - If you'd like to support the MetSci Show financially, here's where you can donate on a one-time or an ongoing basis. Thank you!
- MetSci Show - If you'd like to use our IoT or AI Data Value calculators, or you'd like to contact me, the MetSci Show site is the best way to do it.
- MeteoScientific Console - Use LoRaWAN - The MeteoScientific Console allows you to use LoRaWAN today. As long as you have Helium coverage (and you probably do, about 90% of populated areas in the world have a gateway within 2 miles), you can onboard a sensor. You can always check coverage at https://explorer.helium.com and switch to the "IoT" tab in the top right.
Transcript
Today's guest on
2
:MeteoScientific's The Business of LoRaWAN
is Liliane Paradise, CEO of Miromico
3
:Liliane leads
one of the most technically advanced IoT
4
:hardware teams in Europe, where
she's guiding the development of rugged,
5
:modular gateways and pioneering
ultra low power sensor systems,
6
:some even fully battery free,
from proof of concept mass production.
7
:She oversees the entire arc of building
8
:industrial grade LoRaWAN solutions
with Swiss level precision.
9
:Before stepping into executive leadership,
10
:she earned her chops
as a hands on hardware engineer, helping
11
:build Miromico's earliest LoRa devices
before the LoRaWAN spec even existed.
12
:That depth shows
whether it's pushing energy
13
:harvesting into real world deployments
14
:or designing every product for testability
and manufacturing from day one.
15
:She's making hard technical decisions
with clear business outcomes.
16
:In this conversation, we cover the promise
and tradeoffs of battery free IoT.
17
:The practical realities of hybrid
power systems, and why cybersecurity
18
:and interoperability will drive
the next phase of help you in growth.
19
:We also got into Mioty,
Miromico's manufacturing philosophy,
20
:and how they use early customer feedback
to validate value, not just functionality.
21
:This episode is sponsored by the Helium
Foundation's IoT Working Group.
22
:Helium offers global LoRaWAN coverage
and is used by everyone
23
:from hobbyists to businesses
deploying country wide networks.
24
:If you'd like to see if Helium
coverage exists near you,
25
:check out the links in the show notes
to get started using Helium today.
26
:You can sign up for a console account
27
:with MeteoScientific
at console.meteoscientific.com.
28
:Now let's dig into the conversation
with Lilian Paradise.
29
:Lilian, welcome to the show.
30
:Thanks for coming on. Welcome.
Thanks for the invite.
31
:Super thrilled to be here.
32
:Super psyched to have you on.
33
:I think Robert Bogs had recommended
you come on.
34
:And you and I have been talking
for a hot minute.
35
:Looking into your background, I guess
I've got to start with this obsession
36
:with RF.
37
:Like you're super psyched on radio stuff.
38
:What drives that? Yeah.
39
:So we at Miromico We're actually coming
from a background of IC design.
40
:But from where we branched out to embedded
because our CTO Alex Raymond,
41
:he always had an obsession with RF.
42
:And actually we were doing LoRa before
there was even a LoRaWAN standard.
43
:So we were always really one of the first
to fiddle
44
:around with antennas
and with RF and with new protocols.
45
:For example, one of our first LoRa project
that was even before
46
:there was a LoRaWAN standard, was to build
remote controls for racing drones.
47
:Okay, that was all on LoRa.
48
:I kind of can't imagine
that has a ton of bandwidth.
49
:So those things must just get go left,
go right, go off, go down,
50
:go faster, go slower.
51
:Like, actually,
you could get the latency down if you do.
52
:LoRa.
53
:Just the just the modulation
and not the LoRaWAN protocol on top.
54
:Okay.
55
:Of course if you have LoRaWAN,
you have to delays from the window timing,
56
:but you can get a lot
57
:of more bandwidth through if you just use
the LoRa as a physical modulation.
58
:And the beauty
is compared to other technologies,
59
:you get a hell of a lot of range.
60
:You get all these kilometers
like penetration in a rural environment.
61
:So that's the beauty
about the LoRa modulation.
62
:And it's not tied.
63
:I mean, the LoRaWAN is great, but
the beauty is that it's not tied to it.
64
:You can use it
65
:separately and build really, really cool
solutions based on own protocols.
66
:Yeah. And that's
where we're really, really strong.
67
:Yeah.
68
:And I think we're seeing that
in the kind of wider LoRa world
69
:with the rise of Meshtastic
and then the Loko tracker, stuff
70
:like that,
where people are just saying, hey,
71
:we're just gonna use LoRa
72
:when I'm going to get into the LoRaWAN.
And of course, there's,
73
:you know, different solutions
for different problems.
74
:Is there a favorite LoRaWAN device
that you have?
75
:Is there something that you've worked on
over the years?
76
:You're like, oh, this is awesome.
77
:Well, on one hand,
of course, it's our gateways.
78
:We put a lot of love and effort
and passion in our gateways, worked on
79
:getting them as rugged as possible
and as versatile as possible.
80
:What does that mean in practice?
81
:Our philosophy is always to keep
our products modular.
82
:Modular, meaning that now
we have a LoRaWAN gateway, but we built it
83
:in such a modular way
that it was really easy and simple for us
84
:to build a Mioty gateway,
which is a new, also awesome
85
:LPWAN technology
that came up a few years ago
86
:and is starting to get really popular
and that's not the limit.
87
:We can build a gateway out of it.
88
:We have Wi-Fi support on it,
so that's for sure.
89
:One of the products that I personally
really love and have a passion for.
90
:It might sound a bit weird because
the gateway is always in the background.
91
:It's invisible,
but think about it if the gateway fails,
92
:then you can have all the fancy sensors
you want.
93
:It just doesn't work.
94
:So it's the silent but very,
very necessary part of every network.
95
:And it's the one part that cannot fail,
96
:and the one part that must be secure.
97
:Yeah. Okay.
98
:We'll hit some security stuff later,
because I saw reading up on you
99
:that you're into the security piece.
100
:But let's pause for a second
on this Mioty piece.
101
:I'm hearing a bunch about it.
102
:Obviously it's not LoRaWAN,
but it's important to not kind of ignore
103
:that competition out there
or the other solutions out there.
104
:Give me a perspective on LoRaWAN, Mioty,
kind of what they are doing
105
:that you're excited about.
106
:So the main difference from a
107
:technical perspective is that LoRaWAN,
of course, they use the LoRa modulation,
108
:which gives great range, but they
you basically send out the package
109
:or your whole payload message
as one package.
110
:And when you have an interference,
you lose your whole message.
111
:And that's where Mioty comes in.
112
:They have the so-called TSMA telegram
split in multiple access technology,
113
:where they split the telegrams
into small sub packets
114
:and then they spread them over time,
but also over frequency.
115
:So you're very, very narrow.
116
:You're somewhere around 3
kHz in bandwidth.
117
:And that's just in a probabilistic
calculation.
118
:You're much,
much less likely to have collisions.
119
:Plus on Mioty you also have higher forward
error correction,
120
:which then in turn
means that you can still,
121
:even though it's not that probable,
you can still lose a few of your packets.
122
:Actually, to be technically correct,
you can lose up to 50% of your self
123
:packets
and still restore your original telegram.
124
:And our experiments have shown.
125
:So we have on a local cell tower
in Zurich,
126
:close to our office,
we installed a gateway roughly a year ago.
127
:That's actually almost an anniversary
year, and I have some sensors registered
128
:there and I didn't lose a single packet,
so that's quite impressive.
129
:Yeah. That's pretty.
130
:It reminds me a little bit of Telemacho
was on over at EchoStar.
131
:Was telling me about this. What is it?
132
:LRFSS the long range frequency,
something spread spectrum.
133
:The same kind of thing.
134
:Is that right?
135
:Yes. It goes in a very similar direction.
136
:It's not 1 to 1 the same.
137
:And actually there are some interesting
study studies were conducted by TU Ilminal
138
:that compare the two,
they have quite a similar concept behind.
139
:That's true okay.
140
:Yeah I'm not the most technological guy.
141
:So I guess that's probably
as much as I can understand.
142
:But it's good to hear that
143
:my understanding of it is good enough
that they're similar.
144
:Okay, so that's LoRaWAN, Mioty stuff.
145
:Let's talk a little bit about something
I know you're psyched on, which is this
146
:energy harvesting and the push
either to battery free or hybrid battery.
147
:How are you thinking about that
148
:for LoRaWAN devices
and what are the implications.
149
:Yeah.
150
:So maybe let's start
with a little bit of a background.
151
:During Covid, we started to deploy
large numbers of ambient sensors.
152
:So they measure temperature humidity
but mostly also CO2
153
:in order to ensure the well-being
in offices, hospitals, schools.
154
:And back then, people were thrilled
155
:because we can promise battery
life times up to ten years.
156
:That was five years ago, roughly.
157
:And now, of course, the ten years are,
if you're lucky, with SF7
158
:and now some of the first sensors
that might be in a not so good
159
:location RF wise, far away from a gateway,
they start to deplete and people
160
:start to realize
161
:that it will be a huge operational effort
to replace all these batteries.
162
:And that's one aspect.
163
:And then the second aspect for sure
is the trash
164
:and the toxic waste
that you are producing.
165
:And that's why not just we as a company,
but I think also the European Union,
166
:of course, Switzerland is not part of it,
but we try to move with them.
167
:We'll have directives
on how many batteries you can use
168
:or if you're even allowed
to use primary batteries.
169
:So I think for every IoT company,
170
:it will be critical to bring out energy
harvesting solutions.
171
:And that's what we did.
172
:We already successfully brought our Miro
Insight Lux to the market,
173
:which is completely battery
free and offers all the benefits
174
:our battery powered sensors have so
all the sensors can be supported.
175
:And of course we have a secondary backup
storage there.
176
:We have a supercapacitor,
which is more environmentally friendly
177
:than the primary battery would be,
and it helps to make it through
178
:not just one night, but up to 60 nights
179
:or 60 full days,
to be precise, without recharging.
180
:So that's a that's quite impressive
where technology has led us.
181
:And that's of course,
thanks to protocols like LoRaWAN and Mioty
182
:and also the chip manufacturers,
which really brought down
183
:the energy consumption of the MCU, sent
the sensors really, really low.
184
:Yeah. Interesting.
185
:And then talk me through this idea
of going battery free versus a hybrid.
186
:And what is kind of marketing hype
187
:and what is practical and then where
it might go in the long term.
188
:So as we see on LinkedIn
and basically on every platform,
189
:every magazine
everyone wants to go completely battery
190
:free, which is understandable
from a sustainability perspective.
191
:But in reality, we see, okay,
if the sensor measurement is completely
192
:uncritical and it's a nice to have,
then battery free is the way to go.
193
:But in many cases, the monitoring is more
than just a nice to have.
194
:In some cases,
companies must comply with CO2 levels
195
:or they want to monitor,
for example, the state.
196
:Of course,
that shouldn't be open for too long.
197
:They need to send out alerts.
198
:Let's say in pharma industry
you have a cold room.
199
:If the door is open for too long,
all these medications go bad.
200
:And in that case,
201
:you cannot risk having an outage
because you didn't get light for too long.
202
:And in that case, hybrid systems
will be really, really important.
203
:What does that mean?
204
:A hybrid system means
we still have a primary battery in there,
205
:but we extend the battery life time
by the means of energy harvesting.
206
:Got it, okay.
207
:So it's
208
:this thing where it's like, hey, energy
harvesting and battery free sounds good.
209
:But for right now
let's make sure we have a decent backup.
210
:So if you have a critical sensor, you're
still going to get your information.
211
:Yes, yes.
212
:I think that especially in industrial
settings, it will be critical to have like
213
:these hybrid systems because we're talking
about dirty environment.
214
:We cannot guarantee that the solar cell
or the whatever
215
:the harvesting technology
you're using is always operational.
216
:Yeah, that makes sense.
I can get behind that.
217
:And as fun as the energy harvesting is,
it does make sense.
218
:Like, hey,
219
:let's make sure this thing
continues to work,
220
:even if it has a cool
looking screen on the front.
221
:Yeah, exactly.
222
:All right.
223
:So MIromico,
you've brought a bunch of new stuff.
224
:You kind of watch a bunch of devices go
from concept to, you know, in your hand.
225
:Can you walk me through the arc of what
that looks like
226
:as far as, like, how does that happen?
227
:And what do you have to think of as you're
going through that arc?
228
:Yeah, hundred percent.
229
:So maybe to start with
we are 100% the customer driven company.
230
:So usually an idea starts together
with a vision, a vision
231
:very often from a customer or a partner.
232
:And from that vision we we try to build
233
:a posse proof of concept very fast.
234
:And then once you have a proof of concept,
what we will want to do is to
235
:kind of have a proof of value,
which sounds fancy, but in the end
236
:means we don't need to show anymore
that we can build it.
237
:We want to show that we can actually add
value to our customers.
238
:That means in practice,
we will send out these POC devices
239
:to our Alpha or beta customers
and see what value can they actually gain,
240
:and can we keep the promise
that we're giving to our customers.
241
:So that will be the very first step,
242
:which we basically call technology
and value evaluation.
243
:And from there we usually go through
2 to 3 phases in the product development.
244
:It starts with the so-called MVP.
245
:It's not most valuable person.
246
:It stands for Minimum Viable product,
which is a product that already
247
:looks like a product has,
let's say, 80 to 90% of the functional.
248
:It should, but it's not fully certified.
249
:And with that product,
that minimum viable product,
250
:we will hit the market in a better state.
251
:And we will,
252
:depending on the device, deploy a few
hundreds to thousands of these devices.
253
:And this way we can gather the data
and get the feedback
254
:from our customers
and put it into the final product.
255
:And then the second stage
will be the field testing improvements,
256
:where we do all the pre certifications,
we do all the quality control.
257
:So we ensure that all the environmental
conditions can be dealt with and so on.
258
:And then the output of that phase will be
a ready to be industrialized device.
259
:And industrialization is then
usually the set up for mass production.
260
:Because at Miromico we don't do, “oh
we test one device
261
:and then we ship out the thousand”,
At Miromico we really value our quality.
262
:So every single device
will be end to end tested.
263
:And of course that takes some time
to set up the whole production facilities.
264
:That's
a very Swiss way of doing things like,
265
:we want this thing to be perfect
or it's going to be like a clock.
266
:Yes, it has to be.
267
:That's the promise
we give to our customers.
268
:I mean, there are competitors
on the market that might offer it
269
:for a lower price,
but we really value our quality
270
:and actually a lot of our devices are end
to end, assembled in Switzerland itself.
271
:Wow. That's pretty cool.
272
:Yeah, that's an unusual
it's pretty unusual, I know.
273
:And a part of
274
:it is also social responsibility,
not just for the environment.
275
:Of course, we like to have a value chain
that's concentrated in the region,
276
:but also we work very closely
with a partner that hires people
277
:with certain backgrounds of mental
disability and physical disabilities.
278
:And that way it's a nice social engagement
as well.
279
:And the Swiss
and the Euros are so nice, Americans...
280
:I was thinking
281
:when when you're talking about batteries
like, oh, we have to have batteries,
282
:like America
is going to go find another planet
283
:and the euros are like, hey,
284
:we should actually just stay on this one
to take care of this one.
285
:This one is nice.
286
:We like to why should we trash it?
287
:Yeah, there's. Yeah, there's no good
reason.
288
:Let's see.
289
:So this sounds like a pretty expensive
way of doing things.
290
:What are the biggest cost drivers with
I guess IoT devices
291
:and then IoT deployment.
292
:So I would say on the device level to,
293
:to ensure the quality
that's a big cost driver especially. So.
294
:So the components, the components
you can buy and the components you buy
295
:in very similar conditions
everywhere in the world.
296
:If you have basic set of like negotiation
skills, I would say.
297
:So what really makes the difference is
where do you assemble your devices?
298
:And of course, how efficiently
do you assemble the device.
299
:So as I said,
we assemble our devices in Switzerland.
300
:So in that perspective,
we're more expensive
301
:than if we would do it, maybe in some
other geographical region of the world.
302
:So the way we optimize
is that we try to make the devices
303
:as easy to assemble as possible,
so we cut the time
304
:rather than the employee or stuff cost.
305
:Oh it's interesting.
306
:So you're building with more
than just the end result in mind.
307
:It's also the process is included
in how you're designing it.
308
:I think an important keyword
there is design for testability
309
:and for manufacturability,
310
:which means that we learned every screw
will cost you a few seconds in production.
311
:So whenever you can have a clipping
where you can save the screwdriver,
312
:that already saves
you few seconds the same.
313
:If you can go for a fully automated test
rig that will save you more time than if
314
:someone has to be there and plug something
in and observe a screen sequentially.
315
:And that's where we start in
the very early design phase that we have
316
:these thoughts
how we will industrialize the product,
317
:not just in the very end,
but at the beginning of the design cycle.
318
:It sounds like very, Japanese car
manufacturing philosophy.
319
:It was like,
hey, we're going to think about
320
:this whole thing
and constantly improve it.
321
:The Kaizen stuff and the Deming stuff.
Okay.
322
:Super cool.
323
:Let's see.
324
:Let's finish up on this idea
of what you see
325
:coming down the pipe next in LoRaWAN
that you're super excited about.
326
:And maybe it's LPWAN
327
:and I know you're psyched
in this Mioty stuff,
328
:but what do you see in the next year
to five years that you you're.
329
:Yeah you're pumped on. Yeah.
330
:But on one hand, as you already said,
331
:I think the multi technology
approach will become really big.
332
:Not because.
333
:So we at Miromico
we are a technology provider.
334
:We're not saying LoRaWAN versus
Mioty one is better than the other.
335
:But I think the diversity
will become really really important.
336
:Also, there is no technology
that fits every single use case.
337
:So having a variety of technologies
to choose from
338
:and having them interoperable
will be the next big thing.
339
:So also standardization
between technologies but also between
340
:device vendors will be a really big thing
that will drive deployments.
341
:And we're asking about deployment costs.
342
:And I think that's one big issue
343
:that you buy from different vendors,
and the devices are not interoperable,
344
:and you will spend a lot of time to build
your device models in your back end.
345
:That's that's a huge thing.
346
:And then of course, the other huge thing
that will come up that we are not so happy
347
:about as a device manufacturer,
but that will be super, super important,
348
:is the European Cyber Resilience Act,
which will ensure that
349
:device,
especially gateways, are more secure.
350
:I mean, the whole point of the initiative
is to secure public infrastructure,
351
:and I think that will change
how IoT devices are built, as well.
352
:Do you think that the European
353
:manufacturers will fall behind
in the global marketplace because of that?
354
:Or do you think you'll come out ahead
or what it would
355
:or the implications
for the euros on that side?
356
:I think it will
perhaps push us back a bit for,
357
:let's say, 1 to 2 years,
because let's be honest, cybersecurity is,
358
:if I may say so, a little bit of a pain
in the sense of you need to do
359
:a lot of pen tests and so on and so on
to ensure that you're actually secure.
360
:But cybersecurity is a global topic,
so maybe it will put us behind for a year.
361
:But then once the rest of the world
will come up with similar initiatives,
362
:we will be ahead
and we will have everything ready,
363
:and then it will shift over.
364
:Yeah,
well fingers crossed. Come to California.
365
:We move fast around here here.
366
:We're a little less safe
but we’re pretty fast.
367
:Let's see.
368
:Do you think that AI will just solve
this interoperability problem?
369
:That that's what it seems like
to me is like a non technologist
370
:is we're just going to apply the,
you know, robot brains to this thing and
371
:and that'll fix it.
372
:Or is that wishful thinking here.
373
:I think that the idea behind it
or the direction is smart
374
:or it can solve a huge part of it.
375
:I mean, we already see it with ChatGPT.
376
:I can put in like a bunch of payload
decoder and say, build me
377
:a unified device model.
378
:So it will aid for sure.
379
:But at least from my perspective,
you still need
380
:to be a good prompt engineer
381
:to tell the artificial intelligence
what you want from it.
382
:Okay,
so I think it's part of the solution,
383
:but there still need to be
some brain power, some actual human based
384
:brain power behind
to see what we want to get out of it.
385
:Because otherwise
everyone builds their own standard.
386
:And if everyone has a standard, it's
not a standard, right?
387
:Yeah, yeah, yeah.
388
:No, that still comes down
to, coding is easy, thinking is hard.
389
:That's how the world works.
390
:Exactly. Ripping.
391
:Well, Lillian,
thanks for making the time to come on.
392
:I know you're super busy
running Miromico over there and crushing,
393
:so I appreciate you making the time to
come on. Thank you so much.
394
:That's it for
this episode of The Business of LoRaWAN.
395
:I built this for you.
396
:So whether you're a business owner,
a LoRaWAN professional, or a hobbyist,
397
:the intent is to give you great LoRaWAN
information.
398
:Of course,
the best information doesn't come from me.
399
:It comes from the conversations
400
:we have with the people building
and deploying this tech in the real world.
401
:And that's where you come in.
402
:LoRaWAN is a global
patchwork of talent and ideas.
403
:And ironically,
for a globally connected network,
404
:most of the brilliant folks
working on it are connected yet.
405
:Help me change that.
406
:Introduce me
to someone awesome in your network,
407
:someone doing meaningful work
and laureate work.
408
:Just shoot me an email.
409
:I'll take it from there
410
:and get them on the show
so we can share their work with the world.
411
:You can always find me at MetSci.show.
412
:That's M-E-T-S-C-I dot
413
:S-H-O-W, metsci.show.
414
:If you want to support the show
in other ways, you can subscribe,
415
:leave a review,
share it with your corner of the world.
416
:All those are super helpful.
417
:If you'd like to support financially,
you can go to support.metsci.show,
418
:there are One time and recurring options.
419
:We're also open to sponsors.
420
:If your company serves
the LoRaWAN community
421
:and you want to reach this dedicated
audience, let's talk.
422
:If you want to try out.
423
:LoRaWAN for yourself
create a MeteoScientific account
424
:at console.meteoscientific.com
and get your first 400 DC for free,
425
:which is enough to run a device
sending hourly for about a year.
426
:This show is supported
427
:by a grant from the Helium Foundation
and produced by Gristle King, Inc..
428
:I'm Nik Hawks,
I'll see you in the next show.