IoT technology trends

The opportunity of IoT driven business is to a wide extent based on three tendencies: Cost, capabilities and convenience. Although the cost of sensors, bandwidth and processing has continuously been diminishing and more advanced functionality made possible, the history of technology is littered with countless examples of product introductions which did not live up to expectations set from the announced performance (IBM’s OS/2) or were ahead of time (Apple Newton).
The lessons learnt are to search for a profitable solution to a real problem or desire – not to develop tech gadgets disregarding the customer. The following paragraphs list evolving factors to the perfect storm: trend blurring previous distinctions of technology products; the risk of missing new and unorthodox market entrants; how new possibilities emerge to allow lifelong product improvement to stay relevant; and driven by data insight.

Technology trend: device merger

In 1991 most of the companies delivering the products on this Radio Shack advertisement did not anticipate each other eventually to become potential competitors. However, soon after the mobile network absorbed the phone answering machine. Then the tape/CD player was replaced by an iPod successively succumbing to network streaming services, and apparently, the convenience of having a still camera and video recorder available any time in the pocket is a stronger driver than a high quality SLR camera or camcorder.

The iPhone 7 has more than 1000 times more transistors than the original PC Pentium requiring reciprocally less energy per computational transaction. Heat dissipation per bit of information calculated or transmitted is reduced with every new generation of silicon integration which also contributes to diminishing cost of the components and sensors required to collect and compute data.

Business Trend – delivering on customers’ expectations?

The easy access to almost indefinite computational power poses a threat to the traditional companies with a dedicated purpose and tangible machined solution. New-comers may originate their offer from the abundance of computing virtually ignoring the limitations of the physical world. Hence the existing business owners risk becoming an add-on feature integrated with somebody else’s broader solution, or a mere supplier of goods for an ecosystem where value is extracted from their products.
New market entrants do not necessarily see the existing business vertical as an integrated unit like the current range of well established companies. Newcomers experiment, simplify and carve out what has value or they redefine the service to what customers really want but never got.

Newcomers carving out the ‘secret sauce’

A major contributor to the mobile phone paradigm shift from each-product-its-own-software-variant into the huge App eco systems of iOS and Android releases is attributed to eroding the originally mandatory FCC/ITU (telecom approval bodies) approval for the complete phone including software, menu texts & graphics. The java engine encapsulated a cradle for small separated apps. Now the mobile market is all about the application downloads, while cellular data connectivity remains a small corner of a general computing device as customers may use e.g. WiFi for internet based calls. FCC/ITU has lost control of the entity but still each radio module has its separate certificate proving specification compliance.

Apple’s iPhone redefined use of mobile services

When Apple introduced the iPhone the entire mobile phone market was locked in with carriers/operators demanding full compliance to detailed technical specifications. Moreover the business developers had been accustomed to optimising every new offer to the end user’s cost awareness of high data tariffs in existing mobile business systems. Apple thus created a device with such simplicity and appeal that AT&T was willing to disregard their otherwise mandatory requirements to standards compliance.

Apple further insisted AT&T introduced flat-rate data consumption breaking the existing minimised-data-traffic paradigm by all other product manufacturers. Consequently, the end user experienced a no-frills product stripped from endless menu-options lists and once connected, every application was consistently feeding actual data to/from the network. These factors gave this ‘why has it been so difficult until now’ experience by customers, effectively breaking the gridlock of the existing competition tied to an antiquated data tariffs paradigm.

Medical equipment going digital

A similar development may face medical equipment verticals outside FDA’s traditional regulatory control of the industry. Potentially disruption looms from new business models of sensory gadgets – including sending digital data to an already FDA-approved cloud, compliant to harsh data storage regulations in every relevant market.

R&D Trend: A product’s life starts when unwrapped and powered on for first time

Product development through a stage-gate model delivers a defined feature list at a predictable quality at a certain date as illustrated schematically in the above figure. Following product launch the R&D team ships the maintenance task to an engineering team and thus the development phase is considered completed
Industrial components are traditionally tailor made with exact memory capacity and processor power to fit exactly the purpose. The slim-fit design can be traced back to a time where developing dedicated electronics, every bit and byte represented a considerable expense. Fully customised solutions are often in smaller production series too expensive compared to mass-produced standard COTS (commercial-off-the-shelf) platform offering versatility and full tool chain for fast development and deployment.

Adding more processing power and memory than needed thus opens for continued product development and improvements after shipping as shown in above figure. It is evident a PC comes with extra space for new software capabilities and updates to the operating system, bug fixes and solutions to identified security flaws.
Smaller IoT systems do not normally undergo an aftersales explosion of added functionality similar to PC’s or smartphones but as runtime and market data are gathered. As the numerous examples shown so far indicate, it may be relevant to upgrade the install base with new subscription features and parameters rather than expecting a resell of entire new products. Hence the ability to upgrade and improve the already in-market product may add business value throughout the product lifecycle.
However, once a product enters high volume production, cost optimisation should be introduced as long as changes do not introduce binary breaks – that is, rendering applications and services incompatible after the update.

Siemens Wind Power

The global R&D team has merged with traditional engineering teams overseeing in-market devices: A new title introduced recently as “Head of Research and Development and Product Life Cycle Management” is extending a hitherto sole focus on new product R&D.
The role distinction embraces lifelong value extraction on running Wind Turbines monitoring 3.200 individual measured values each on 7.800 turbines and identifying parameters for continuous R&D effort and improvements.

Pattern Recognition Trend: Knowledge from multiple devices – machine learning

Initial data collection will not necessarily require deep insight into computing and cloud functionality. Accumulating data will eventually become an asset and its importance will stretch beyond the actual business opportunity in the making. Since experimental and standardised data collecting units are commercially available, setting up experimental data connectivity to the network may be arranged with an industrial cloud provider. Subsequently analytics tools are getting still more user friendly for visualising the findings from data stored in the cloud.
A potentially valuable proposition is to search for patterns in the utilisation of a large population of behaviour data, then sharing the benefit of insight to everyone using the service. Manufacturers collecting run-time data systematically across multiple installations can foresee apparently uncorrelated events predicting faults appearing in turn. The capabilities of advanced statistical algorithms applied across massive data sets will reveal dependencies not identified by usual monitoring and reporting.

Google Nest

The Google-owned NEST thermostat is learning from monitoring daily habits and accommodating for persons indoor or approaching the house for adjusting heating in advance. Google gains valuable insight in how customers behave at home while adding safety of smoke- and CO-detecting sensors. The electronics in NEST is over-specified for the rather simple task of controlling the heating system as defined in the first product releases – recently the system has been integrated with the voice operated Google Assistant.

Moreover the NEST is controlling a vast array of heating systems already in the market which constitutes a threat to eliminate the specific user intimacy of existing competitors – or an opportunity to build a strong b2b asset where Google feeds machine learning usage data and product specific knowledge back to individual suppliers for optimisation of hitherto independent heating systems.


Continue reading: go to Business Innovation Theme