IoT - Why IT Must Fail Fast to Succeed

The Internet of Things (IoT) revolution will START, SLOW, STOP then SUCCEED!

Taking us into cycles of hype before throwing us off at breakneck speed onto a harder landing than when we boarded this ride.

I'm not going to focus on the Things in IoT. My case for IoT is really a business model for Things as a Service (TaaS). In short, Stop Buying, Start Paying for Things, but I get ahead of myself... let's find out why?

START? I've had the great fortune to participate in an evolution of ideas, discussing IoT with a home grown innovator of social objects in Singapore. I love the hearing about his journey from 2006-2010 getting his patent approved, today he stands at the forefront with early adopters.

IoT somewhat means that the internet is involved, it's not for example the handheld digital pet, a Tamagotchi that can communicate peer to peer using infra-red when another Tamagotchi is nearby, does not require the internet. IoT will have objects using the internet, an example is your smartphone when you allow it to use your current location to recommend good restaurants rated by your friends nearby.

My view of IoT here is tempered by looking past smart objects to social objects towards Things as a Service.

SLOW? The IoT-ing (i o t -ing) of complex objects to make them smarter and connected is not fairly hard with the low tech sensors to collect perpetual data streams or high tech circuit boards to process and communicate actions to take. Even simple objects can be made smart with enough ideas for their use at hand.

An object is just a thing, a smart object is able to connect online to a base and takes it's instructions from it. A social object is a thing that can become a friend of another thing (trusted to take actions from an exchange of instructions between them).

Unfortunately we have to first slow down and ask why would we even want this to be possible? For this to happen, we'd be changing paradigms, thinking of things talking to other things and asking what if they could talk? what would they say? what would they do? This will keep us busy incubating ideas and concepts then some time building working prototypes.

STOP? Bringing an object to market will involve packaging it better than the IoT or TaaS name we've given them, the mass market does NOT care if it has sensors, can collect data, use an internet medium or peer to peer communication. Folks only care about purchasing Things that work for People and naming the tech part of it so that its as well recognized as WiFi isn't easy. Yet that's just the easier part.

The problem part presents for sophisticated objects when the object itself costs more than the normal commodity, like the Herman Miller Sayl chair that I've pictured above. Unless we're creating addons for objects which can limit their design, the acceptable threshold that folks are going to buy it, is limited by their past experience buying similar objects.... that leads us to ask ourselves, when a sensor is turned on and requires an internet connection to stream data somewhere, who will foot that bill?

TaaS is more than IoT, where a social object will need an always online connection to make new friends of social objects, sometimes permanent, sometimes temporary so that both objects can exchange data to check if an action must/can be taken (eg WeChat service). BUT to be really useful that cloud exchange would need to do some analytics ie crunching data to understand it fast enough to know what instructions to send. This can be done peer to peer with enough compute power but I'm looking at scenarios where it's most cost effective to analyze in the cloud. The best example is like buying an appliance that requires electricity to power up and paying to use that power supply whenever the appliance is used.

As long as TaaS requires an always on connection to the internet which could be your home broadband or tethered to your cell phone when out but since they are also social objects someone must pay for it to stream data to be analyzed somewhere on the cloud exchange so that preset instructions between them may be sent at appropriate thresholds to take an action.

IF THIS: I've been sitting with the same weight distribution on my Miller Herman Sayl for 3 hours, which can happen when I'm composing long emails, and my chair (a social object) told my laptop (it's social object friend) that I've had enough.

THEN THAT: my laptop goes into un-wake-able sleep mode for an hour to remind me to get up and about. This can prevent those really annoying times where my left thigh shoots a sharp pain up my nerves often when I've sat for too long. My pre-agreement with my wife is for me to get out to the gym at the clubhouse.

While this communication can be established via Bluetooth, my pair of social objects also notifies my car which is always parked outside the clubhouse to turn on the car alarm if I wasn't out in 10 mins to deactivate it before going to the gym.

And just for fun if I didn't reach a heart rate of 130 on the threadmill before deactivating my car 45 mins later, the threadmill or car sends a WhatsApp message to my wife to prompt her to check up on me. She just wants be to be really inconvenienced if I cut short good gym.

For that to happen, the sensors would have needed to be timing me from the moment I sat down and be streaming that data or holding that data till the right time to initiate those instructions. Granted not much data crunching is going on, but you get the idea when you put into play a limitless number of social objects :)

Unless broadband is nationalized or globalized and some governing body collected an internet tax to grant unlimited broadband access to all and any who would want to use it, really like a highway.

As long as the model to pay for the data to be streamed over the internet is uneconomical or unwieldy, the adoption of TaaS and I daresay for many of the early IoT which will have an "offline" feature will be halted. Remember Folks only care about purchasing Things that work for People, and most don't care how it works... imagine having to contend with paying for an internet

SUCCEED? Having lived thru an early internet hype cycle for Application Service Providers to Hosting to Virtualization to Cloud, I can say that vendors that went out of business 10 years too early selling software as a service, helped create the necessary awareness for hosted and cloud services market today. Today any CIO worth their salt wants to consider what the cloud can do for their business.

To re-start the IoT momentum a few things may happen. The model below is just my humble opinion, of which came of working with both the hosting and cloud utility business models.

What If? Things were really sold as a service and there wasn't a huge upfront cost to "BUY" it.

What If? We "PAY" for Things over the course of it's useful lifetime, like a subscription.

What If? A TaaS business was setup with the capital outlay to acquire all the "Things" we want to use for IoT and "sold" them to us like a lease or hire purchase agreement.

A TaaS business could be very successful with a Telco type company. They could start by creating a menu of essential social objects that have a healthcare need before moving on to necessities or luxuries. What social object first is a longer topic that I'll not cover but here's a comparison of why I think these types of business can enable and grow the IoT and TaaS industry

A Telco already have the capability to provide the social object as a bundled Thing as a Service much like what they already do with cell phone/router bundles for their contracted subscriptions. Likewise they can both be charging for and collecting payments for the ISP service as well as the cloud exchange for data crunching. It's just a stretch today to expect them to bundle TaaS necessities or luxuries that may cost a lot more than the devices they are used to giving away for free as a cost of acquisition or paid over time.

That's where partnering with a bank or finance house that can establish credit worthiness before granting loans and collecting installment payments over the loan period to acquire that TaaS. Installments could factor in the cost of streaming data over the internet and making the same regular payments to the Telco/ISP without inconveniencing the user of the social object.

Other businesses, like leasing, hire purchase or energy utilities also could provide a headstart to the TaaS market, having common experiences. In the case of utilities they are the best example of company who collects a fee for the usage of any electrical object.

Bottomline, the business model needs to change from one where folks purchase everything upfront and never pay for using it ever to one where our mindset also changes to one where we don't ever pay for any thing upfront but continue to pay for using it forever (this already works for expensive objects eg a car, now think like my Miller Herman Sayl chair).

This post is made while I'm a Microsoft employee and the opinions and views expressed here are those of my own and does not necessarily state or reflect those of Microsoft Singapore or Microsoft Corporation.

If you value this post, pls feel free to ‘like' or share it with your LinkedIn, Facebook, Twitter or Google+ friends. Comments are highly appreciated. Thks for reading! See my other posts at https://www.dhirubhai.net/today/author/davidtang