Global IoT Trends Report - 2022
Executive Summary
The annual survey of the State of IoT conducted by element14 showed some interesting results in 2022. Companies are engaging in IoT development because they see opportunities to grow their business and establish leadership in their industries, but can’t do it alone. The research showed a growing trend towards working with partners to develop solutions, and a demand for standardisation and interoperability, rather than companies working on their own to develop closed systems.
IoT continues to be deployed in a wide variety of applications. This year, Industry 4.0 remained the primary application, representing the primary market for 18% of the respondents, and we also saw very strong growth in energy management from 10% in 2021 to 16% this year. The increase in energy prices around the globe after the survey closed suggests that energy management could be even more important in 2023.
This year, there was a dramatic change in the attitude to deploying AI in IoT systems. In 2022, engineers are better informed and more willing to deploy AI as they find ways to use it to improve products and increase the RoI of their systems.
There was a slight trend towards custom boards, rather than off-the-shelf SBCs, which could be linked to an increase in the deployment of low-power wireless communications. We also saw a wide variety of sensors being deployed, reflecting a broad range of applications. The huge diversity of IoT applications demonstrates the need for support from vendors, such as element14, that offer a wide portfolio of products – from SBCs to MPUs and wireless communications modules to sensors – so engineers can quickly and easily source the best products for their requirements.
Methodology
The survey was conducted from January to March 2022. Responses were solicited from people involved in IoT development using a variety of channels, including social media and email. An incentive was offered to encourage participation.
A total of 2445 responses were received. Respondents who completed fewer than five questions, entered text into open questions that was not relevant, or who appeared to be bots or duplicate entries were removed, leaving a total of 2263 valid responses.
Margin of Error
There is no consistent margin of error as questions were not made compulsory, allowing respondents to skip questions, and partially complete surveys were included in the analysis. Assuming a population of about 4 million and 2263 responses produces a margin of error of +/-2% at 95% confidence level.
Survey Demographics
Most respondents were in technical roles, with 52% engineers and 5% technicians. Some of the other categories are likely to include technical respondents, particularly the owner/manager category, where for SMEs this role is often held by a technical person.
There were a significant number of non-technical respondents. One potential improvement in future iterations of the survey would be to ask some different questions of the less-technical respondents.
The survey reached a global audience, but is not representative of the distribution of engineers or engineering projects, as it is skewed towards respondents in Europe and North America.
Gender demographic
With a technical audience, the survey received the vast majority of responses from males(91%). This reflects the need to attract more women into engineering to ensure the industry benefits from diversity. The small percentage of women (6%) makes it impossible to produce reliable analysis of the difference in response by gender.
Overall the demographics of the respondents was like that for the previous year’s survey, and so we can compare the results from year to year without being concerned that the samples are sufficiently different to skew the results.
What’s Driving IoT?
What do you believe will be the top IoT industry in 5 years’ time?
The respondents didn’t have a clear view of the industries that are likely to drive the growth of IoT. When asked what they thought would the biggest IoT market in five years, the responses were distributed evenly across the different sectors.
Industry 4.0 was identified by 18% of the respondents, followed by energy management at 16%, home automation at 15% and artificial intelligence at 14%. Note that the difference is not statistically significant between adjacent markets.
When compared to the results from previous years, it is clear that energy management is becoming a more important application (growing from 10% to 16% in this question). It is interesting that the research was complete before the large energy price rises. It seems reasonable to assume that energy management will become even more important as the cost of energy increases in the short to medium term.
There were a small number of respondents that cited other industries, particularly agriculture and environmental monitoring. Although only a small number cited these other industries, it might be worth considering including them in the survey next year as there will obviously be a bias to the industries that are listed as options.
Which application do you think will grow in future as a result of the impact of COVID-19 on IoT?
The pandemic has had a major impact on technology. When asked which applications would grow the most due to COVID, health-related applications (healthcare and wearables) accounted for 40% of the respondents’ answers.
Remote asset access was the next most likely to grow, with 19% identifying this as a market that would expand because of the pandemic, and the other markets followed behind with similar percentages.
Overall these results reflected the responses in the previous year, although robotics/drones showed a significant increase from 6% to 9%. It’s not clear why this has happened, and should be monitored closely in future years.
In your opinion, what is the reason for the slow growth of Industry 4.0 despite advancement in IIoT technologies?
Industry 4.0 is an interesting case as it has not grown as fast as some predicted, despite the level of innovation and development in industrial IoT (IIoT). The respondents were very divided over the reason for the slow growth of industry 4.0, citing technology limitations and the difficulty of retrofitting more commonly that other potential causes of delay. Factors such as RoI, lack of skilled workforce and concerns over data security, however, were also identified as major factors in the disappointing rollout of industry 4.0. It’s interesting to note that suppliers like element14 are actively addressing many of the issues: for example increasing their portfolio of industry 4.0 suppliers and providing content and resources to enhance knowledge and skills. It could be that this investment by element14 and other suppliers boost the deployment of industry 4.0 over the coming months and years.
This question was asked differently in previous years, moving to a single select in 2022. The technology limitations/difficult to retrofit option was added this year, and instantly became the primary reason cited for poor growth of industry 4.0. This suggests that there is a great opportunity for companies to develop solutions that simplify the deployment of IoT on brownfield sites.
Interestingly the next second cited problem was RoI in 2022, whereas this was the fourth most common answer in previous year. It could be that the RoI from industry 4.0 deployments is becoming harder to identify, but equally could reflect the technical issues having been solved and companies focussing on financial criteria to decide if they will deploy industry 4.0. It also shows that there are still market opportunities to develop cost-effective industry 4.0 solutions that offer more compelling RoI than today’s products.
Factors Influencing IoT Rollout
Out of the following, what do you feel is the most important aspect to consider when developing IoT solutions?
The survey asked what the biggest concerns and considerations around IoT for the respondents were. In both cases security topped the list, with communications/connectivity second. This shows that there are still fundamental challenges around IoT to be resolved if the technology is to build on its initial success.
The ecosystem was another aspect identified by respondents, showing that companies don’t feel confident that they can create IoT solutions without partners and support from other organisations. It would be interesting to conduct further research to understand exactly what the concerns are, and whether they apply to particular parts of the ecosystem.
Reliability of the edge device fell significantly as a factor to consider from previous years, dropping form 15% in 2021 to 10%. This suggests a maturity to IoT, with edge devices now assumed to offer adequate reliability by most respondents.
What is your key concern regarding IoT implementation?
Interoperability was the third biggest concern identified. This goes a little further beyond the worries about ecosystem and shows that the respondents feel that companies need to avoid the development of closed systems. This is something that is also reflected in other questions, as well as being similarly important in the previous year’s results. It would be interesting to understand more about the perceived benefits of developing an open system vs the lock-in created by closed systems.
Which of the following factors is most likely to accelerate adoption of the IoT?
When asked about the things that would accelerate IoT openness was the clear winner. Open standards were cited by 22% of the respondents, interoperability/certification of standards by 17% and 13% identified connectivity standards as being the most important factor. Relative to other factors, ease of development fell compared to the results in 2021, suggesting that it might now be more straightforward to develop IoT solutions than previously.
There was an increase in the important of the public sector in the Far East, with 23% of Chinese respondents citing this as the primary factor to accelerate IoT, which was generally reflected across the Far East. Open standards were much less important, with ease of development the second-most important factor in China. Interestingly, however, 15% of respondents in the USA – significantly above the proportion across the whole sample – also cited public sector involvement as the key factor.
IoT Development is a Team Effort
Do you use a third party for any part of your IoT systems design?
Around half of respondents (52%) use third parties, something that has not changed significantly since last year. Third parties were used mostly for designing sensors and nodes (20%) – this was a new category added to the research this year, and so we cannot compare the areas where third parties help with previous studies.
The use of the cloud is responsible for driving companies to seek partners, with 13% using others to do edge to cloud communications and 10% seeking other organisations for data centre and analytics support.
How likely is your company to take on the lead role of building IoT solutions?
The importance of an ecosystem of companies is illustrated by the fact that one third (34%) never take the lead role in development, and the same percentage only sometimes take the lead role. This means that the majority of the projects undertaking by the respondents’ organisations are not led by them.
Although the fact that companies are not leading could be seen as a negative, it also indicates that there are strong relationships between organisations. In fact, we believe it’s more likely that companies could be choosing to operate in an area in which they specialise, rather than trying to offer an integrated solution that might require expertise that they don’t have. Whatever the reality, partnerships and collaboration are crucial for most companies developing today’s IoT solutions.
What is the main reason for your company developing an IoT solution?
When asked why they develop IoT solutions, not leading the project does not mean that companies are not innovative. Although the main reason for developing IoT solutions was a practical “because there is a need in the market” 27%, one quarter of the respondents said that they did so because their company was an innovator. A further 13% worked for organisations that had made IoT development a business objective or focus.
Only 15% of respondents said that they were developing IoT products to follow an industry trend. This, coupled with the previous responses about taking the lead role, suggest that most organisations are developing IoT solutions to lead their market, but the complexity involved requires that they build an ecosystem of partners.
Privacy and Data Ownership
The survey showed that there was considerable concern around data privacy and ownership. 64% thought that the user whose data was gathered should own that data, and 7% felt that external governing bodies should control the data. Less than one third (29%) thought that the company who gathered the data should own it.
Although these results are in line with the findings in the previous year, it does highlight the need for companies to consider who owns the data that is generated from the IoT. However, we suspect that although the individuals working at the companies feel that users should control the way their data is used, commercial realities mean that a smaller percentage of deployed systems give the users the ownership of the data.
This mismatch between the systems that are being deployed and users’ expectations raise the possibility of companies gaining a competitive advantage in the IoT market by introducing systems that enable the user to continue to own their data. By taking such an approach, companies could build trust with users, which is likely to help them grow market share.
The Value of IoT
Where would businesses gain the greatest value from the IoT Data?
The primary value of IoT was productivity and manufacturing improvement, cited by 30% of people. This is perhaps not surprising, as the rollout of IoT in the industrial sector has been faster than other sectors such as consumer. Other benefits were mixed, and with the huge variety of IoT applications now launched, perhaps benefits are now more application-specific than attributable to IoT as a whole. This reflects the growing maturity and diversity of deployment associated with IoT technologies.
IoT Software
What programming language do you use for your IoT development?
The languages used for IoT development remain more or less unchanged from last year. C/C++ (44%) is the most popular, followed by Python (28%) and then JavaScript (17%). Several respondents used more than one language in their development.
Other languages used include C#, Rust (a language that is very similar to C++, but which guarantees memory safety), and Squirrel (a lightweight scripting language).
Communications for IoT
Please choose your preferred type of wired connectivity
Communications/connectivity was identified as the second-biggest concern for IoT development, so is an extremely important factor to consider. The vast majority of respondents (76%) use wireless communications, leaving 24% using wired connectivity.
Ethernet dominated the wired connections, with 69% using Ethernet or EtherCAT. There was some use of industrial standards such as CAN (10%), PROFIBUS (5%) and ModBUS (5%). Interestingly there was also some use of industrial low-level serial connections, with 10% of respondents using RS232 or RS485.
With USB becoming widely used in many applications, it was interesting to note that this was not an option offered to the respondents and no one mentioned USB when citing other wired connectivity standards.
Please choose your preferred type of wireless connectivity
The situation was slightly different for wireless communications, with three standards achieving significant usage: WiFi (45%), Cellular (20%) and BLE (15%). Presumably these represent applications that have a need for connectivity over moderate, long and short distances respectively. The proportion of respondents choosing these standards probably reflects their application requirements more than any particular preference.
LoRa is the connectivity standard used by 12% of the respondents. This is interesting, as the market for LPWAN is still developing, and certainly not as mature as WiFi or Bluetooth. It was also noticeable that SigFox lags a long way behind LoRa, with only 2% of respondents choosing this LPWAN standard.
Of the respondents using other standards, ZigBee/802.15.4 was by far the most popular choice. Although, perhaps, less fashionable than it once was, it would be worth considering including this as an option in future research.
Despite the many highly developed standards, there were a few people using Sub-GHz (4%). This shows that companies are still prepared to invest in proprietary communications links. Presumably they are not the same organisations looking for interoperability, which is more likely to be enabled by standards such as WiFi.
It’s interesting that there is a significant variety of wireless solutions for IoT connectivity, and even in the wired domain there is a need for alternatives to Ethernet. It is obvious that relying on one approach is a poor solution, unless the deployments are highly standardised. Distributors with a wide portfolio of communications products, such as element14, enable engineers to equip their IoT nodes with the optimal communications standard for each situation.
Growing Use of Artificial Intelligence
Proportion of Respondents Using AI
The proportion of respondents using AI has jumped from 39% to 56% in a single year. Additionally there are a further 27% of respondents who are considering using AI in the future. This suggests that the technology has become more accessible and easier to deploy, as the change is quite dramatic.
What is the primary application of Artificial Intelligence in your IoT design?
Looking at the answers about AI more carefully reveals an interesting story. Preventative maintenance is the most popular use for AI (16%), but robotics (12%), image classification (11%) and business analytics (10%) are all also popular uses for the technology. This variety of uses for AI shows that no one application or requirement is driving the adoption of the technology, rather there are multiple factors that are causing IoT developers to deploy it.
Equally impressive is the reduction of respondents who are not convinced by the benefits, or do not understand AI. People who thought that AI was not useful for their application dropped from 32% in 2021 to 13% in 2022. Similarly, only 3% said they didn’t understand AI in 2022, a large fall from the 13% who didn’t understand the technology last year.
In the last year we have seen suppliers such as element14 redouble their efforts to support engineers who are considering deploying AI in their systems. The growth of development platforms and SBCs and off-the-shelf AI software, together with educational resources, seem to have eliminated delays in AI deployment due to lack of skills or knowledge.
The responses to the question about AI show an explosion in interest and deployment of the technology. As more engineers understand the technology and its potential, the benefits are being realised by companies who are quickly joining the AI bandwagon. Companies that are ignoring AI risk being left behind.
Hardware for IoT
What hardware platforms do you use to design your IoT solution?
SBCs and customer hardware dominate the platforms used for IoT. In the last year there has been a growth in the percentage of people using custom hardware, and it now matches those using SBCs. This could be, in part, due to the systems being designed. We saw low-power wireless technologies being widely deployed, which would typically be associated with custom hardware, rather than an off-the-shelf SBC. Manufacturers’ development boards are also used by 19% of respondents.
Please choose your preferred Single Board Computer for your design
Raspberry Pi (46% and Arduino (26%) dominated the SBC category, with some indication that Arduino might be taking a small amount of market share from Raspberry Pi. Beagleboard decreases slightly, and Avnet products showed a large jump in usage, although the survey might skew the results towards the Avnet platform, as it was run by a company owned by Avnet.
Please choose your preferred development platform vendor for your design
The survey revealed that engineers were using a wide variety of solutions for their IoT boards, which is probably driven by the diversity of requirements and applications. Partnering with a vendor that can offer the leading SBCs off-the-shelf, as well as offering components to enable the development of custom boards is the best way to ensure that the right solution is selected every time. With Arduino, Raspberry Pi, Beagle Board and a myriad of components, element14 is an ideal distributor to allow engineers to optimise their solutions every time.
STM (28%) and Microchip (23%) each represented more than double any other vendor’s market share when it came to silicon vendor boards being used in IoT. STMicroelectronics, however, has seen a fall in market shar by 9 percentage points, while Microchip has grown from 21% to 23%.
Other vendors cited by respondents included Expressif, Nordic Semiconductor, Silicon Labs and Texas Instruments. Of these, Expressif had the largest share, representing about 3% of respondents. In future research it would be interesting to include these four vendors as options to get a better understanding of this fragmented market.
Please choose your preferred MCU/MPU vendor for your custom hardware
When developing customer hardware, STMicroelectronics (34%) and Microchip (30%) are even more dominant than when choosing a manufacturer’s development board. As with the development boards, Microchip has made some inroads into STMicroelectronics’ market share, growing from 28% to 30% while STM shrank from 36% to 34%.
The same four other vendors were most frequently cited in the other category - Expressif, Nordic Semiconductor, Silicon Labs and Texas Instruments – but they represented a much lower percentage of the total respondents building custom boards. Of these vendors, Nordic Semiconductor was the most popular choice.
Like most other elements of the hardware platform, there is significant diversity in the choice of processor architecture for IoT applications. With no one vendor dominating the market, it seems that engineers are selecting processors based upon specific characteristics that are required for a particular application. element14 offers a vast range of processors, from chips to MCUs and MPUs on development boards and SBCs. It’s interesting that element14 also sees engineers selecting different options based upon the project requirements, and offers a varied portfolio to ensure they can meet the needs of any system.
Cloud Services
As was the case last year, the respondents split fairly equally between the big three cloud providers (AWS, Google and Azure) and using a personal solution. In 2021, however, 34% used a personal solution, and this dropped to only 26% in 2021. This suggests that the benefits of “rolling your own” cloud service are decreasing as the commercial services offerings for IoT mature. We would expect to see this continuing as developing a cloud solution is not trivial, and therefore requires a good reason to invest in such a project.
Since 2021, Google has overtaken AWS as the leading choice of commercial cloud provider (although the margin of error means that this lead could be very small). While all cloud services are investing to win IoT business, Google has been very active over the past couple of years, and this work could be bearing fruit as it builds share in what is a very competitive market.
Despite the dominance of the three big cloud vendors (and custom-builds), there is room for smaller players in the market. IBM Watson has maintained a 3% share of the market, and in the other category a wide range of different vendors were cited. These included hosting companies such as Digital Ocean and OVH; telcos such as KPN and Orange; specialist IoT services such as ThingSpeak and AdaFruit IO; and services such as Ali Cloud that was used by some Chinese respondents.
A variety of Sensors
As with many of the questions in the survey, asking about the sensors being used revealed a broad diversity of applications. 87% of the respondents used sensors in their application, with environmental sensing by far the most popular (34%). Opto/image sensors and motion sensors each accounted for 15%, location 10%, health 8% and audio 5%. 2022 was the first year that we asked this question, so it will be interesting to see how the responses change over time. It would also be interesting to ask about the number of sensors per system in future research.
One of the key findings of this survey was that there was huge diversity across IoT applications. This is particularly seen in the sensors deployed. It is clear that engineers not only need to find a supplier with a range of sensor products, but because different suppliers specialise in sensing different things, engineers also need to partner with a supplier that offers a wide range of manufacturers. It is not surprising that many of the respondents choose to work with element14, as their broad and diverse range of products allows them to select the optimal approach and components for every IoT application.
Useful Links:
Sensors: