Episode 5

From Handfuls to Hundreds - Bansi Talks Scale

Bansi from Macnman joins us to unpack the intricacies of LoRaWAN and its significance for scaling solutions tailored for the Indian customer. We dive deep into how LoRaWAN can transform various industries by enabling the deployment of numerous low-power, battery-operated sensors that collect invaluable data. Bansi shares insightful use cases, particularly in smart agriculture and smart city applications, emphasizing the need for mass deployments to truly see a return on investment. He also highlights the critical balance between the technology's strengths and its limitations, especially regarding time-sensitive applications.

Takeaways:

  • Deploying hundreds of sensors is crucial for maximizing ROI with LoRaWAN technology.
  • For IoT applications, understanding the strengths and limitations of LoRaWAN is essential.
  • Automation in industries like agriculture can significantly benefit from LoRaWAN's long-range capabilities.
  • LoRaWAN is not suitable for time-critical applications that require immediate response times.
  • Effective use of LoRaWAN involves collecting meaningful data over a large network of devices.

Links referenced in this episode:

Companies mentioned in this episode:

  • Macnman Technologies
  • Quicklogic
  • Helium Foundation
  • Meteoscientific

💲Support the show - https://support.metsci.show

Transcript
Speaker:

Welcome to

2

:

the business of LoRaWAN,

where we dive into this long range,

3

:

low power, wide area network

and its impact on your bottom line,

4

:

the latest sensors and proven real world

solutions.

5

:

I'm your host, Nik Hawks.

6

:

Since 2020,

when I stumbled into LoRaWAN via Helium

7

:

while hunting for a lost paraglider,

I've helped thousands understand

8

:

how it works and how to use it,

and I'm psyched to do the same for you.

9

:

I've installed everything

from soil sensors

10

:

to weather stations to retail foot

traffic counters.

11

:

Coming to the conclusion

that LoRaWAN is rad,

12

:

so I built the only show on the internet

fully dedicated to unlocking its potential

13

:

for your business.

14

:

Let's dig in.

15

:

Joining us today is Bansi co-founder

over at Macnman Technologies,

16

:

an IoT hardware company

specializing in LoRaWAN solutions.

17

:

As a director at Macnman ,

Bansii has been instrumental in designing

18

:

and deploying LoRaWAN gateways

19

:

and sensor nodes that power long range,

low power IoT networks.

20

:

With expertise in industrial automation,

PCB design and embedded systems.

21

:

He has led projects that bring real world

applications of LoRaWAN

22

:

to industries like smart agriculture,

logistics and health care.

23

:

Under his leadership,

Macnman has collaborated

24

:

with global tech firms like Quick Logic

to develop customized IoT solutions,

25

:

and their products are used

26

:

by companies needing robust and scalable

wireless connectivity.

27

:

Today,

we're diving into the business of LoRaWAN

28

:

with Bansi to explore

how Macnman is shaping the future of IoT.

29

:

Bansi, thanks so much for coming on.

30

:

Yeah, thanks for having me.

31

:

I'm super excited.

32

:

So I saw that done.

I think it was LinkedIn.

33

:

You had posted an article

34

:

something like the seven Reasons

a Business Should Use LoRaWAN,

35

:

and I thought,

oh, it's your perfect to come on,

36

:

let's give folks who are watching

a little bit of a background

37

:

on what Macnman does

and your role in Macnman .

38

:

Yeah, sure.

39

:

I'm a cofounder of Macnman ,

so we do mostly a core deep

40

:

technical stuff, core and also

connectivity be part of the solution.

41

:

So we are not an end to end solution

provider company, but we do hand out

42

:

connectivity in the non

cellular in the transaction basically.

43

:

So our expertise lies in the battery

operated devices.

44

:

That is our major core strength and

the business perspective I do believe in

45

:

like deploying a so many devices

that collect a very small piece of data.

46

:

And ultimately it will form a meaningful

picture in front of the business owners.

47

:

So that's what we help with that.

48

:

Industries. Got it.

49

:

And you guys have a LoRaWAN gateway right.

50

:

That you guys make yourself.

51

:

Yes.

52

:

We do have under Raven Gateway. Yes.

53

:

Okay.

54

:

Would you walk me through a business

use case where LoRaWAN is going

55

:

to be an excellent fit?

56

:

So I like

57

:

to give it from the example

what we have done in the past.

58

:

So one thing is we've done

an automation in the greenhouse.

59

:

So just getting a data to that

like in the case of the Indian farmer

60

:

is not, sufficient. Right.

61

:

We were pushing some data to the farmer

in the mobile phone,

62

:

and we are sending some notifications

and all stuff, but it is not enough for,

63

:

in order

to bring a real value to the solution.

64

:

So we developed an controllers.

65

:

So these are working on the durum.

66

:

And so based on the algorithm

we deployed I think 40 plus sensors

67

:

in one acre of land.

68

:

And based on that like very small

pieces of data, we were getting

69

:

a meaningful picture of what

is actually happening in the greenhouses.

70

:

And based on the conditions,

we were controlling, the humidifiers,

71

:

dehumidifiers and fats.

72

:

So I think as such, use cases

have a lot of potential.

73

:

Very wanted to, deploy sensors or

you wanted to, go for a mass deployment.

74

:

So LoRaWAN becomes very handy

when it comes to the scalable solutions.

75

:

If you are planning to deploy it

for a very small number of devices,

76

:

I think where in between 510 devices,

it is really difficult to get ROI.

77

:

From my experience, you need to go for

mass, at least for a few hundred devices.

78

:

Oh, interesting.

79

:

Okay, so if you're only deploying 5

to 10 devices, there,

80

:

there may be better other solutions.

81

:

Is that right?

82

:

Yeah. Yeah, yeah, I do think it does.

83

:

It is right by,

due to the reason LoRaWAN requires

84

:

an initial hefty amount of,

infrastructure cost at the initial stage.

85

:

So in order to compensate that cost, you

need to deploy a large amount of sensors,

86

:

which can operate on a battery for, let's

say, five years, six years, ten years.

87

:

And you will be continuously

getting a small amount of data.

88

:

Okay.

89

:

And then you said maybe 100 sensors.

90

:

Is there a minimum number of sensors

you think people should be deploying?

91

:

It is not a minimum or it is not a

I need a rule something like this.

92

:

But that is I think, quite sweet number

which we found in the Indian scenarios

93

:

and Asia-Pacific scenario that as the few

hundred devices are like good enough to,

94

:

see the ROI, like in a one year

or two years after deployment.

95

:

Got it.

96

:

And so

I think I saw, a gateway on the Mccammond

97

:

site was, I don't know, about 350 USD.

98

:

And then what are the costs of the sensors

that you guys are deploying for?

99

:

The cost of the sensor varies

a lot depending on the budget.

100

:

So one question we recently launched is

that has an, multiple inputs.

101

:

So any end user doesn't really

102

:

have to rely on us

to like format custom PCB for himself.

103

:

He can trigger of hardware.

104

:

And according to his requirements.

105

:

So the end solution will be tailored

for his and use case.

106

:

So that sensor starts

with somewhere around

107

:

3,540 USD in India.

108

:

Okay, so if they're getting 100 of those,

let's say for grand

109

:

for the sensors and then they put

in, I don't know, two gateways usually.

110

:

Yeah. Or are you putting in

multiple way more than two.

111

:

No, no the two gateways

are like more than five.

112

:

Yeah.

113

:

Okay. So you got a primary and a backup.

114

:

And so you're looking at 40

I don't know 640.

115

:

Yeah 4600 bucks for a deployment

at least as far as hardware goes.

116

:

Yes. As buy the hardware hardware goes.

117

:

Yes. Got it.

118

:

And then was I reading correctly

that you guys have a LoRaWAN to Modbus?

119

:

Bridge? Basically, yes.

120

:

Yes, yes.

121

:

That is

that is kind of a hard sell product.

122

:

So, that device actually allows,

123

:

a whole different set of new features

and new set of, like sensors to work with.

124

:

Right?

125

:

So, yeah, a few applications.

126

:

One of them is obviously,

underground water level monitoring.

127

:

So there are industries,

as has already installed some sensors,

128

:

and they wanted to just collect data

129

:

in order to have and like that is kind of

first step under the digitalization.

130

:

It becomes very handy

to have such converters because that

131

:

drastically reduce the cost of like,

reinstalling all whole new sensors.

132

:

And you can use existing infrastructure

in better and efficient way, I think.

133

:

So that product is yeah, we do have it

in a lineup that is actually super cool.

134

:

Let's see when you're looking at

businesses, is most of your business

135

:

dealing with agriculture

or are there other sectors that you also

136

:

that management

also deals with, or is it totally wide?

137

:

No. Most

138

:

of the clients are the industries

actually.

139

:

Okay.

140

:

Agricultural is kind of a secondary

a third number client

141

:

and the second one is a smart city.

142

:

So we were doing some early flood

warning systems.

143

:

These are basically ultrasonic sensors

which measure the water level.

144

:

So that is one place called as Mumbai.

145

:

It does near Pune.

146

:

So, so that has me kind of

a critical complex geographical area.

147

:

So if a rain of two hours like it,

it does easily to

148

:

flood whole city.

149

:

So we were generating some early warnings

for the floods.

150

:

Oh super cool.

151

:

And in that case,

the city is really looking at

152

:

mitigating or avoiding the cost of damage

that might happen

153

:

if they didn't know about a flood.

154

:

So yes, how do you,

when you're selling that to the city,

155

:

how do you kind of position

that as a value piece?

156

:

Because they're not instantly

saving money.

157

:

You're just basically saying, hey, you,

158

:

you should be able to save some lives

in the future.

159

:

I'll tell you, when I gave me short

the interesting story about it.

160

:

So the particular Mumbai city and,

all the city,

161

:

there are some environmentalist

who have an, logic,

162

:

of the water level in the particular area

is, let's say x, meter.

163

:

So they do have an algorithm

like after 3 or 4 hours,

164

:

the water in some crucial areas

will be at X plus one level.

165

:

So they do have some algorithm for it.

166

:

So it is in a very good way.

167

:

So they can have an R, they can do

an hourly arrangement for the people,

168

:

for the vehicles, for the, to save

a lot more like future problems.

169

:

So they have they can see in the future.

170

:

But the technology.

171

:

Okay.

172

:

Cool.

173

:

So for them on the city side,

basically the this

174

:

the super simple selling point is like

this allows you to see into the future

175

:

and make sure that you don't sustain

massive damage.

176

:

Okay, cool.

177

:

Is there something else

where you go into a business?

178

:

Because this podcast is all business

of LoRaWAN, and you've got a very clear

179

:

value proposition for them like, hey,

you spend $10,000 on this or whatever

180

:

the number is, here's

how fast it's going to get back to you.

181

:

Can you walk me through a use case

like that or.

182

:

It's very clear.

183

:

Yes, sure.

184

:

So one use case which we recently

done is in the fire pump station.

185

:

So that is in the BMS sector actually.

186

:

So you know building

buildings have the five components added.

187

:

So one interesting use case building

to be installed a lot of controllers

188

:

and install a lot of sensors.

189

:

More like fire

detection sensors in the building.

190

:

So at one point,

191

:

it happened

actually somewhere about a year ago.

192

:

So that is a

193

:

small case of fire

in the in a particular, room.

194

:

So due to the controllers.

195

:

So nobody has to run to switch

on the controllers in the pump room.

196

:

So it all started automatically,

and it saved a lot of money

197

:

based on the automation they have done.

198

:

And the controllers does not cost

that much at some very cost around,

199

:

in the same range,

to:

200

:

So they installed somewhere around

60 or 70 controllers I think. Yep.

201

:

So for the pumps and the sprinklers.

202

:

So the pumps started automatically due

to the indication of the fire in the room.

203

:

And in that way they saved up like,

204

:

very critical and crucial stuff

for in the rooms as well.

205

:

And ultimately the people's live as well.

206

:

So that is

I think, the most like valuable thing.

207

:

What we have achieved till date.

208

:

I like the saving human life,

but I do believe in,

209

:

just sending a notification

or just, gathering a data is not enough.

210

:

You need to put up your value to it

unless.

211

:

And until you do it.

212

:

That the IoT doesn't make any sense

either.

213

:

You need to do a meaningful

analysis of it,

214

:

or either

you need to control the applications.

215

:

I think the real power of IoT lies

in the control.

216

:

You got it?

217

:

Yeah, I know, Thomas Overton

cart has always told me that

218

:

nobody looks at dashboards

that you have to actually use the thing.

219

:

So it's really good to hear you

confirming that.

220

:

Let's see, one thing

I heard at this recent conference, I was

221

:

that was that, LoRaWAN was too slow

for a couple of businesses

222

:

is walk me through where and I'm assuming

with controlling you're using class C,

223

:

so it you know, to me

it doesn't seem like it is that slow,

224

:

but could you walk me through

kind of where it's

225

:

really appropriate to use it

and maybe where it's not?

226

:

If someone's listening to this

and thinking about, oh, should I use

227

:

LoRaWAN in my business

were a great ways to use it,

228

:

and then ways where it's

like, oh, it's not for that.

229

:

You know, the

230

:

protocol has been built in a not on time

critical sense, I think.

231

:

So the protocol is built in that way.

232

:

So it is not built for a very, very time

critical applications.

233

:

So you need to look for n

for another technology.

234

:

If you have any time sensitive data,

it cannot

235

:

communicate like in a few milliseconds.

236

:

It is not a time critical protocol.

237

:

So some of the businesses

that I can suggest is,

238

:

the application

like a smart city, applications

239

:

where you can deploy a lot of sensors

240

:

and you need a really cost effective,

solution or cost effective.

241

:

So the, technology to work with.

242

:

And that should be enough, right?

243

:

So one gateway

can easily add up to:

244

:

So that is where actually

the real value brings like actual value.

245

:

And some of the industries that I think,

246

:

Laura mentioned beyond it will come

something users

247

:

want automation embedded

for the LPG refilling plant.

248

:

So these are

these applications are very time

249

:

sensitive and time critical applications.

250

:

They need to collect and control the data

every milliseconds.

251

:

So that is also even not possible

with the glossy.

252

:

So I do not recommend a time

critical application just the duration.

253

:

But right.

254

:

If you integrate

is not going to change with

255

:

let's say 30 or 40s of delay,

256

:

then it should be a van due

257

:

to like taking consideration of factors

like a scalability and the cost

258

:

effectiveness of the solutions

and the sensors.

259

:

Got it. Okay.

260

:

So the general kind of idea for someone

who may not be familiar with LoRaWAN

261

:

is that if your application can handle it,

what is it, a 30 to 40 second

262

:

delay,

then LoRaWAN is totally fine for that.

263

:

And if it's less than that, you know,

then as you get closer and closer

264

:

to kind of the five or or 10 or 1 second,

it's just less and less appropriate.

265

:

Okay. Yes.

266

:

So for fire suppression, it's fine

because 30s it's not, you know, fantastic.

267

:

But it's also 30s is pretty fast

compared to a human running down the hall.

268

:

And, you know, turning something.

Yes. Much faster.

269

:

That. Okay.

270

:

Very cool. Yes, actually.

271

:

And in the fire station, we,

the power was available.

272

:

So the sensors are not battery operated.

273

:

So that is first thing

I want to all our work in the glossy.

274

:

So and the glossy

you get a little bit faster responses.

275

:

Yeah. Yeah.

276

:

They're they're listening all the time.

277

:

So it should be a little bit quicker okay.

278

:

Very cool.

279

:

Let's wrap this up.

280

:

Is there anything

281

:

any advice that you have to a business

that's considering a LoRaWAN deployment.

282

:

What should they look for.

What should they know about it.

283

:

What would really help them

kind of make a good decision.

284

:

Much faster than otherwise.

285

:

Yeah, sure.

286

:

The real advice I like to grab is

287

:

you need to collect a meaningful data

first time, and you need to.

288

:

You can split your structure,

like infrastructure, in such a way

289

:

that you can deploy a lot of sensors

with the, like, battery operated devices

290

:

and, a few hundred units

so that you can have

291

:

a better image

of what is happening on your shop floor,

292

:

what is happening

in your agricultural field,

293

:

installing a 1 or 2 sensors is not enough.

294

:

In that case, you're not,

compensated any infrastructure cost

295

:

on on one sensor on a one gateway,

you need to go for 100 units.

296

:

I think.

297

:

And the second is,

298

:

again, it would be, kind of same point.

299

:

It does not build for time like time

consecutive applications

300

:

in a particular way.

301

:

So it is not a time sensitive protocol.

302

:

So we need to take care of

and you need to understand

303

:

what are the limitations and what are the

powers of the actual technology.

304

:

What brings to the like real world.

305

:

And you do have a better idea

of your end application.

306

:

Then you can select

a particular IoT protocol for it.

307

:

Got it right.

308

:

One protocol is not enough to cover like,

all the use cases.

309

:

That is for sure.

310

:

You need to have a combination

of protocols.

311

:

You need to work on multiple technologies

if you want it to bring a real value

312

:

to the solution.

313

:

Got it?

314

:

Okay, Bunty, this has been super useful.

315

:

Thanks so much for making the time.

316

:

Thanks for coming on early my morning.

317

:

Probably late in your, in your day

and sharing your knowledge with us.

318

:

Yeah. Thank thanks for having me.

319

:

And it was really nice.

Nice to talk to you.

320

:

Have a good day ahead. By.

321

:

That's it for the business of LoRaWAN.

322

:

Thanks for listening.

323

:

If you enjoyed the show and want to learn

more, the podcast home on the web is meet

324

:

that's.

325

:

Net Sky dot show.

326

:

There you'll find calculators to estimate

the impact of IoT usage on your business.

327

:

Be able to make guest suggestions.

328

:

If you know someone who you think

329

:

should come on the show

and easily get in touch with me.

330

:

If you think the show is useful

for LoRaWAN, please leave a review

331

:

wherever you listen to this.

332

:

Ratings

interviews really help podcast grow.

333

:

Finally, an enormous

thanks to our sponsor, the IoT

334

:

working Group at the Helion Foundation,

for supporting this show.

335

:

If you want to try LoRaWAN out

for yourself

336

:

without asking anyone permission,

you can sign up for a video

337

:

scientific account

that console Dot Meadows scientific.com

338

:

and get your first 400 data credits

for free.

339

:

That's enough to run a sensor for

about a year if you're sensing once a day.

340

:

Right on. See you on the next show.

Next Episode All Episodes Previous Episode

About the Podcast

Show artwork for The Business of LoRaWAN
The Business of LoRaWAN
Learn From the Pros

Listen for free

About your host

Profile picture for Nik Hawks

Nik Hawks

Incurably curious, to stormy nights and the wine-dark sea!