IoB Insiders Analysys Mason analyst Tom Rebbeck says the industry still needs to make IoT adoption easier for end-users deploying these technologies.
The Internet of Things is not living up to expectations. Sold on the idea of a massive new growth market, executives are starting to get frustrated at the slow pace of adoption. The press has also picked up on this, with articles on why the Internet of Things is failing to take off. One recent piece said: “Consumers, developers and even venture firms are becoming disillusioned in building connected hardware.”
How justified is this negativity? And if IoT growth is slow, what can those involved in IoT do?
Other than IoT specialists, like Sierra Wireless or Telit, few companies report IoT revenues. In itself, this is telling; for most firms involved in IoT, associated revenues are too small to warrant a separate reporting line. Where they are reported, the numbers are typically low – four percent of revenues for Intel, one point three percent for Vodafone. Revenue ramp up is rapid but not explosive, with annual growth rates at around 15 percent on average.
The drivers to IoT are well documented
The drivers for Internet of Things growth are well understood and widely discussed – the increased performance and reduced cost (and size) of technology mean that connectivity can help to lower costs and make new revenue opportunities possible. Governments have also been instrumental in pushing for connected solutions, from smart utility meters to e-Call legislation. Finally, we have a strong push from suppliers looking for new markets.
Barriers to adoption merit more attention
Barriers to IoT are less well explored. Solving these problems will help open the IoT market and may create opportunities in themselves.
The first barrier is uncertain demand for IoT solutions. Especially for consumer IoT services such as smart home solutions, demand remains uncertain and is only slowly emerging. The apparent success of standalone devices, like Amazon’s Echo, that require no integration (unlike, say, a Nest thermostat) may point the way here. Initial products may need to be extremely simple, even if that limits some of their functionality.
A second issue is that of fragmented standards. Multiple standards and platforms exist in each part of the IoT value chain. Organizations may be withholding investment until an agreed set of standards emerges.
Organizations may also be developing new standards, as existing standards are viewed as sub-optimal. We believe this is a mistake – as Andy Stanford-Clark of IBM, and one of the developers of the MQTT protocol, says “Don’t, under any circumstance, develop your own protocol [for IoT]!”, as plenty already exist. Even if existing solutions are imperfect, developers may be better off working within their constraints rather than trying to build support for yet another option.
More can be done to illustrate the business case
The third inhibitor is the business case. The business case for many current IoT use cases centres on relatively simple cost savings. For example, for Hitachi Construction Machinery, GE and ThyssenKrupp – three organizations that have made early moves in connecting devices – their marketing centers on how adding connectivity reduces the operating cost of an expensive machine.
There are fewer examples of business cases involving reduced costs for less expensive devices, or of IoT generating new revenues. We believe that vendors can do more here – plenty of proof points on the investment case for IoT exist – but more could be done to provide case studies and even example models to illustrate the investment case.
Privacy and security concerns are the fourth major barrier: Enterprises and end users are unclear about the privacy and security implications of billions of connected devices, and with some justification.
A recent paper “Don’t Panic: Making progress on the ‘Going Dark’ debate” listed examples of how IoT devices could be used for government surveillance (and we’ve recently reported on NSA activity here – Ed), stating that “the future will be even more laden with sensors that can be commandeered for law enforcement surveillance”.
The implications of this are concerning – if devices can be hacked by governments, who else could gain access? We see a lot of work being done here by vendors to show how security is incorporated into solution design, something that is increasingly being demanded by customers. The IoT director of a telecoms operator recently told us “Two years ago security in an IoT RFP was a ‘nice to have’, today it’s a ‘must-have’.”
A fifth barrier is around data discoverability. Major IoT solutions, for example, for smart cities or intelligent transport, need resolution of issues such as data discovery, data formats, payment for data and so on. One study estimated that less than 15 percent of organisations with IoT solutions integrate them with third-party products and services.
For the Internet of Things to realize its potential, solutions may need to take actions based on data from many sources: one estimate is that 40 percent of the value of IoT will be from solutions using multiple sources.
At present there are many barriers that make using data from multiple sources difficult or impossible, and this can limit the impact of IoT. Solutions are needed to the knotty problems of making data discoverable, available in usable formats and easily purchased. Several companies and organizations are working on resolving this – both from a standards perspective (with projects such as FIWARE and HyperCat), to create marketplaces to discover data (Onetransport, datastreamx, Thingful) and ways of paying for IoT data (Tilepay). Solving this barrier could develop into a lucrative business. Paul Egan of 8Power (formerly with IoTUK and Neul) has predicted that “the first company to come up with a data exchange for IoT will probably be the next Google.”
Finally, while governments can be catalysts for investment, they sometimes create unnecessary uncertainty. Any uncertainty about regulatory and government policy positioning may delay or defer Internet of Things investment. We know of one example where a telecoms operator was planning to launch a wide scale LPWA network, but halted the plan after the government suggested that it could build a national LPWA network and that any public-sector initiative would use this government network. 18 months after the telecoms operators’ initial plan, there is still no sign of the government network. Governments and regulators need to be clear in any statements relating to the Internet of Things and understand the potential unintended consequences.
The inhibitors to IoT can seem daunting – the points covered above are a long (and incomplete) list and none has a simple solution. By their very nature, all IoT projects will be complex. Solutions will involve multiple elements – sensors, connectivity, data storage and processing, software and so on. Even simple services are hard and sometimes painful to develop. Added to this point is the lifetime and replacement cycle for many devices. ‘Things’, from thermostats to industrial machines, will last many years and it will take time for investment to filter through. IoT firms need to be realistic in their short-term expectations but should also try to identify how they can reduce barriers to adoption.
Tom Rebbeck is the lead of Analysys Mason’s Digital Economy research practice