A look at data sources

There are several stages of IoT data collection, depending on the use case; the types, origins, and properties of data can differ significantly.

For example, when considering IoT in smart offices, the goal is usually to improve workplace productivity, safety and security. Energy monitoring and saving devices, such as a commercial smart thermostat, tend to play a prominent role in such deployments, and the data they collect is often located in offices, rooms conference rooms or specific floors. Comprehensive analysis software may not be necessary in such settings, and indeed the quality, cost, and fidelity of the devices themselves become more important concerns.

In contrast, industrial sectors, including factories, are likely to have clusters of devices in one location, and the data is also multi-granular and heterogeneous.

Other examples include retail, where data can be collected from weight sensors on shelves, facilitating inventory control and providing real-time alerts, and healthcare, where even wheelchairs are equipped with IoT sensors.

Across industries, there are creative uses for devices — and the data they collect — that are likely unknown at this point.

IoT analysis software is required

To take full advantage of the data generated in this environment, as well as to enable truly profitable optimizations and efficiencies, businesses should invest in IoT analytics software with the following features:

• East ready for business. The software should come with all the necessary features to ensure that the organization can facilitate a complete IoT data analysis operation. Ideally, the offer is available in SaaS mode and provides the guarantees of confidentiality, security and customer support essential to the operation and marketing of business software.
• Is flexible in terms of devices. The IoT analytics offering should be able to ingest data from a variety of devices and not be overly constrained or specialized given how quickly the IoT is growing and evolving. There is a natural trade-off between a software’s specialization for one industry or company and its ability to adapt to other industries. Leaders in this space must be able to effectively navigate these trade-offs to retain their position. Startups in the IoT analytics space should explore underserved niches in this sector, but eventually need to address scaling issues.
• incorporates machine learning. While much basic analytics, and even visualization, can be done without machine learning, integrating it into the analytics pipeline is essential for organizations to achieve greater efficiency and optimization. Analytics software that makes machine learning transparent has a competitive edge.
• Uses advanced models, such as Explainable AI. Machine learning and AI are not the same thing, although they are often confused. Explainable AI that can build on – but is distinct from – machine learning will be important for IT teams who don’t want a black box of machine learning models that tells them what to do, but instead provides reasonable explanations underlying any recommendations.

  The incorporation of explainable AI will allow startups and small businesses to enter the IoT analytics market and charge a premium for their services. Potentially, companies offering explainable AI specializing in IoT could partner with more traditional IoT analytics vendors that have scale advantages in software testing, delivery, and marketing.

• Offers dashboard, visualization and reporting. Visualization has always been an important and often underestimated component of analytics pipelines. A good platform should support visualization and the products that are a direct result of it. More advanced software options allow operators to quickly and easily generate visualization and data-driven reports and help business managers make decisions in short cycles.

Future prospects for IoT data collection and analysis

An exciting industry trend is the push towards greater standardization of connected devices which include industrial IoT, smart home devices, and internet-connected devices. Such normalization is inevitable in the next two to three years. Due to this trend, analytics software is expected to become cheaper and more interoperable over the next few years.

Executives may decide to wait to invest in expensive software in the near future only to find that it is superseded by a superior option. Instead, the next year might be better spent planning, budgeting, strategy, and training. For a business heavily reliant on IoT, waiting may not be an option, but executives should be wary of longer-term contracts for existing software offerings.