New York City’s infamous jail, Rikers Island, currently houses nearly 7,000 detainees. Chefs, not inmates, do the cooking. But while they’re on their shift, the chefs are locked in, too. There are cameras everywhere, monitored from the guard’s office. Knives are chained to heavy machinery. Can lids slide into a locked cage. Spoons are locked up in the office. Inside Rikers’ kitchens, there’s a delicate balance between the chefs, guards, and the detainees who wash the dishes.

• My son became quieter as he entered his tween years.
• Car rides turned into our most meaningful time together.
• Showing up consistently matters more than forcing conversations.

When my son was still under 5, I worked at the local library. In my free time, my son was basically my best buddy. My supervisor at the time had teenagers, so she was in a different parenting stage than I was.

I remember her saying that the secret to keeping her kids close was to drive them around as much as possible. This kept her kids talking to her and enabled her to maintain close relationships with them.

She was so right with her advice.

My son is now a tween

My son just turned 12, and the shift from boy to teenager seems to have happened overnight. He is becoming more independent and less talkative — with me, but not his social circle.

I know it’s the natural order of things for him to spread his wings and to push back a bit against me. But sometimes I just really miss my best buddy and all the fun we had spending time together when he was little. Now, when he’s home from school, he’s often in his room talking on the phone with his friends or playing games online with them.

I think my former boss’s wisdom stuck with me, because the idea of my son growing up and not wanting to talk to me scared me. I’ve come to realize that driving my son around whenever I get the chance is basically priceless. Right now, drives with him are primarily to and from school, but during football or basketball seasons, all of the practices and games really add up. This year, sixth grade has really felt like a turning point, because I’ve noticed an increase in invites to parties, hangouts, and sleepovers.

I realize that as he gets older, these social outings will only increase. And then one day, when he’s closer to age 16, he’ll likely have a part-time job to add to his schedule. As long as he doesn’t have a car, I know I’ll be his main source of transportation. Instead of dreading, I know these are actually the hidden opportunities, like diamonds in the rough, to remain connected to him as he grows up.

It’s best to allow our conversations to flow naturally

I never try to force a topic on him, because I have found that it’s not the best timing for discipline-based or serious talks. I’m sure he feels trapped, so he shuts down, and it ruins the safe space I’m trying to develop out of our car rides. Allowing the conversation to flow organically is when he’ll surprise me and ask me something random or open up about something that’s been bothering him.

Even if he doesn’t open up every time, I know I’m giving him the space to do so. Often, after a few minutes of being stuck in the car together, one of us will start talking about something. I think having the music on and sightseeing on our way everywhere gives our brains distractions and talking points. It feels like the car is sometimes the white flag zone, where we stop arguing and start talking again.

While he’s mostly reserved, there are other times when he’s more open and chatty, and I just let him vent and do my best to listen. It’s likely therapeutic to have someone who will just listen to him at his age, but it might also be easier for him to open up to me side-by-side instead of face-to-face. Knowing there’s an endpoint, such as knowing we’ll be at his school in five minutes, likely helps too.

I hope I’m also sending him the message that I won’t stop showing up

Willingly taking him everywhere he needs to go daily, I think, is communicating to him that I’m not going to stop showing up for him. That no matter how tense things may be at times between us, I’m going to continue to be there for all of it.

I think it reassures him that I’m not going to give up on my job as his mom, even when things get tough. I’ll be sitting there in silence if that’s what he needs, but the message I hope to send him is: I’m still here.

• Trump says he will impose a 15% ‘worldwide’ tariff.
• He earlier said he would impose a 10% tariff.
• The Supreme Court struck down many of his tariffs on Friday.

President Donald Trump is not giving up on his tariff strategy.

The president said Saturday in a post on Truth Social that he would now impose a 15% ‘worldwide tariff.’

“I, as President of the United States of America, will be, effective immediately, raising the 10% Worldwide Tariff on Countries, many of which have been ‘ripping’ the U.S. off for decades, without retribution (until I came along!), to the fully allowed, and legally tested, 15% level,” he wrote.

In a 6-3 decision on Friday, the Supreme Court said Trump did not have the authority to impose his tariffs under the International Emergency Economic Powers Act, a national security law that allows the president to regulate economic activity during emergencies.

In a press conference following the decision, Trump said he would use a separate authority to impose a 10% global tariff on top of any existing tariffs.

The separate authority is known as Section 122, which can only be imposed for 150 days. After that, Congress must vote to extend.

A White House official later said that countries being tariffed under the authority the Supreme Court struck down will now be subject to that 10% tariff.

“With IEEPA no longer applicable, those countries will now be tariffed at the global 10% tariff using the Section 122 legal authority,” the official said in a statement to Business Insider. “This is, however, only temporary as the Administration will be pursuing other legal authorities to implement more appropriate or pre-negotiated tariff rates.”

The president’s Truth Social post on Saturday, while short on details, indicated he would raise that tariff to 15%.

The IEEPA-justified tariffs have been among Trump’s most powerful weapons in his efforts to renegotiate trade agreements worldwide. They included Trump’s so-called “Liberation Day” tariffs, announced in April, which are at least 10% on nearly every country in the world.

“During the next short number of months, the Trump Administration will determine and issue the new and legally permissible Tariffs, which will continue our extraordinarily successful process of Making America Great Again,” the president wrote in his Truth Social post on Saturday.

The IoT platform market is mature enough to have dozens of options and immature enough that many of them aren’t really platforms at all. Choosing the wrong one doesn’t usually blow up at deployment. It surfaces later, when margins shrink, projects stall, and switching costs are already too high.

After years of watching system integrators, OEMs, and enterprise teams navigate this choice (and sometimes deeply regret it), I’ve distilled a 20-point selection checklist. But before you go through all twenty, here are five traps that keep destroying projects and business relationships. These are not edge cases. They are patterns.

Trap 1: You’re Evaluating a Dashboard Builder, Not a Platform

A vendor shows you a slick demo. Nice charts, clean UI, data flowing from an MQTT broker into a time-series database, some alerts on top. Looks like an IoT platform, right? It’s not. A dashboard builder with an MQTT broker bolted on is not an IoT platform. A device management tool with a web interface isn’t one either. And a single-vertical solution repackaged as something “generic” is probably the worst offender: it looks convincing until you try to do something outside that one vertical.

Before you evaluate anything else, make sure the platform covers the absolute minimum: multi-protocol connectivity (not only MQTT/HTTP), bi-directional device control, structured hierarchical data modeling, event-driven processing with correlations and root cause analysis, native visualization, API-based integrations, and a security model that goes beyond TLS and passwords.

If even one of these is missing, it’s not an enterprise IoT platform. It’s a component that will require five more components around it, each with its own integration cost, maintenance burden, and point of failure. Walk away early.

Trap 2: “MQTT and REST” Is Not a Connectivity Strategy

This one is deceptively simple. MQTT and REST work beautifully for modern sensors and clean APIs. The first project flows, everyone is happy. Then comes the second project with industrial PLCs, Modbus, OPC UA, and even legacy serial protocols. The platform can’t connect to half the devices, and the team is stuck.

Even if you’re building just one solution today, think about tomorrow. Your next customer or vertical will almost certainly bring devices that don’t speak MQTT. And if every unsupported device requires an expensive gateway or the vendor’s professional services to integrate, your margins will erode with every new project.

What you actually need is broad industrial protocol support covering both IT and OT, an SDK or driver framework you can use yourself, and the ability to build custom protocol adaptors without calling the vendor every time. Gateway and edge connectivity should be part of the deal, not a separate product with a separate price tag. If the platform locks you into “MQTT/REST only”, you’re selecting a constraint, not a platform.

Trap 3: Weak Data Modeling: the Silent Margin Killer

Data modeling is invisible during demos and tends to surface only when real money is already on the table.

Most platforms offer some form of “device twins” or “digital twins”, essentially a flat structure where each device has properties and maybe some metadata. That works for monitoring dashboards. It does not work for real IoT solutions where you need hierarchy: Enterprise → Factory → Workshop → Production Line → Unit, or Building → Floor → Zone → Room → Sensor. These aren’t cosmetic layers. They define how data flows, how access control works, how alerts propagate, and how operators actually interact with the system.

Ask explicitly: does the platform support a formal, platform-wide data model? Can you visually design custom business object hierarchies, not just flat device lists? Can you define parameters, operations, event types, and dynamic object-to-object bindings? Is this reusable across projects?

Here’s a practical test I always recommend: take the platform’s demo and try to build your actual asset hierarchy. Not the vendor’s example. Yours. If it takes more than a day and still feels awkward, you have your answer.

When data modeling is weak, every project turns into custom development. You hardcode what should be configurable, duplicate what should be inherited, and every new customer multiplies the problem instead of reusing the solution. This is how SIs end up losing money on projects that looked profitable at the proposal stage.

Trap 4: Cloud-Only Deployment is a Time Bomb

Cloud-native is great. Cloud-only is a serious strategic risk, especially in industrial IoT, telco, and government sectors. A significant share of enterprise customers either can’t or won’t put their operational data into a public cloud. Utility companies have data residency requirements. Manufacturers want everything behind their firewall. Telcos need to run the platform in their own data centers. This isn’t a niche concern; it’s the reality for most industrial deployments.

If the platform only runs in the vendor’s cloud or supports just one public cloud provider, you’ve put a ceiling on your customer base. And the vendor’s promise of “we’ll add on-prem support later” should not give you any comfort. Cloud-to-on-prem portability is an architectural decision that either exists from day one or costs a fortune to retrofit.

The checklist is rather obvious: on-premise deployment, private cloud, multiple public clouds (not just AWS or Azure), hybrid architectures, cloud-provider agnostic operation, and edge deployment. Mandatory for any serious industrial, telco, or MSP-oriented business.

Trap 5: The Edge-Cloud Frankenstein

This trap is particularly nasty because it hides behind good-looking slides. An “edge solution” and a “cloud platform” that work together seamlessly. On paper. In reality, these are often two separate codebases — sometimes from two separate acquisitions — duct-taped together via APIs.

The practical consequence is painful: what’s built for the cloud gets rebuilt for the edge. The team maintains two platforms instead of one, needs two sets of skills, and runs two deployment pipelines. Code portability between edge and cloud remains a slide-deck fantasy rather than an engineering reality.

What you need is fundamentally simple (and surprisingly rare): same codebase across edge and central platform, same development tools, no “rewrite for edge” requirement. When a model, a dashboard, or an alert rule designed in the cloud can be deployed to an edge node without modification — that’s real edge-cloud compatibility. Everything else will cost you months per project.

Ask the vendor directly: “Is your edge product the same software as your cloud product?” If the answer involves phrases like “lightweight version” or “seamless integration between our two products”, you know what you’re dealing with.

These Five Are Just the Start

Platform selection has more dimensions than connectivity and deployment. Think vendor maturity, AI readiness (yes, it’s 2026 and MCP servers and development agents are part of the conversation now), security architecture, pricing transparency, and the uncomfortable question of what happens if the vendor disappears.

The key takeaway: platform choice defines your profitability, scalability, and reputation for years to come. Weak platforms have a way of silently destroying margins before anyone notices.

I’ve put together a comprehensive IoT Platform Selection Checklist covering the full picture. It’s designed as a practical yes/no/maybe tool: if too many boxes stay unchecked, walk away.

→ Read the full IoT Platform Selection Checklist (2026 Edition)

The post IoT Platform Selection: Five Traps to Avoid appeared first on IoT Business News.

• Sarah Canzano is the mother of Gavin Canzano, a 15-year-old aerial skier.
• She and her husband recognized their son’s talent early on.
• They have been supportive of him in his quest to qualify for the US Ski Team and the 2034 Olympics.

This as-told-to essay is based on a conversation with Sarah Canzano, the mother of Gavin Canzano. It has been edited for length and clarity.

I’m the mom of an elite teenage athlete. My husband was a downhill skier as a teenager, and he taught our boys, Gavin and Deacon, to ski when they were about two or three. I don’t ski at all.

Gavin was a shy and reserved kid, but he was a different kid on the mountain. He was so confident. We used to say that skiing was like oxygen to him. It’s where he is at his best.

It was easy to recognize his talent early

Gavin began skiing with the Bristol Mountain Freestyle team when he was around 8 years old. It’s a feeder into the US Ski Team. It was apparent quickly that he was very good in the air. In 2024, when he was 14, he was invited to Lake Placid, N.Y., to participate in the US Ski Team’s Project Gold, an aerial camp. He was all in on aerial skiing after that.

Now 15, he is part of the US Ski Team’s national development program for aerials. He trains with his national teammates at the Olympic Training Center in Lake Placid. He also travels for competitions.

It can be challenging to raise a teenage athlete and create balance

He’s a 15-year-old boy. He makes mistakes. But as long as he’s working hard, and doing the best he can at school and being a good human, we will continue to make sacrifices for him.

We have also had conversations with Gavin about the sacrifices that he would have to make on this path. That means school is going to be harder, because he’s going to miss a lot of school. He won’t be able to see his school friends that much when he’s competing and traveling. But, we’ve always told him, “We will work hard for you as long as you are working hard for yourself.”

He loves what he does, which makes it easy for us to support him. If he ever got to the point where he didn’t want to do it anymore, we would never make him. It sounds insane considering the investment we’ve made, but it’s his body, his life. We are also teaching him to advocate for himself because we won’t always be with him.

We’ve had to change our lifestyle to accommodate him

I’m a travel agent, and we like to travel as a family. Before, we’d be off to an island somewhere every break, but now I’m spending that time standing on a mountain. We’ve definitely had to change our plans several times for him, but we never put that burden on him. This is our family’s choice.

My husband and I divide and conquer a lot. A lot of times, just my husband or just me will travel with Gavin, so that Deacon can be home, in school, and with his friends, pursuing his hobbies. Pat and I are apart a lot, but we also try to take advantage of that one-on-one time with our kids, and the moments we have together with them.

Deacon is the real hero here; he is such a supportive little brother. He goes along for the ride with minimal complaints and is Gavin’s biggest fan. He’s so proud of him.

We’re taking it one step at a time, but the Olympics are the goal

We want him to have fun. He had me write “have fun” on his skis, so if he’s having a bad day or a couple of bad jumps, he can remind himself that this is fun.

But it’s also serious. We might consider looking for financial sponsors soon, as we try to get him to where he needs to be. We will cross those bridges as they come. The goal is the US Ski Team for the 2034 Olympics. That sometimes feels overwhelming — it’s almost a decade away.

My job as his mom is balance. I have to let go of some control, but I also need to remember that he’s 15 and still needs a lot of guidance from us. I am confident in his abilities, and I’m now able to watch him jump without closing my eyes anymore.

Pecorino Romano is one of the oldest cheeses in the world, with roots going back to Ancient Rome. But today, most of it is no longer made near Rome at all. In this episode, we visit I Buonatavola, one of the very last producers still making Pecorino Romano in Lazio, the cheese’s original territory, to understand how global demand, especially from the United States, reshaped where and how this cheese is made. We explore the differences between Pecorino Romano made in Rome and the versions produced elsewhere and how exports helped keep this historic producer alive.