The purpose of this article is to present the actual applications of combining AI and IoT being used in industry to generate strong value, and to demonstrate important trends that will create even more compelling future use cases.
So, there is a clear connection (no pun) between the Industrial Internet of Things (IIoT) and Artificial Intelligence (AI).
If IoT is about connecting (usually wirelessly) machines and making use of the data generated from those machines (via visualisation platforms or App), AI is about adding an extra software layer, using that data to bring more intelligent behaviours in machines of all kinds.
IoT needs AI and vice versa: As IoT devices will generate large amounts of data from different sources and type of sensors, then AI will be functionally necessary to deal with these huge volumes to make more sense of that data. Data is only useful if it creates an action. With IoT, you can geolocate & monitor equipment, you can receive alerts if a sensor data is outside of thresholds, you can receive meter index…. But to make data truly actionable, it needs to be supplemented with context. IoT and AI together is this context, i.e. ‘connected intelligence’ and not just connected devices.
Traditional methods of analysing structured data and creating action are not designed to efficiently process the vast amounts of real-time data that stream from IoT devices. This is where AI-based analysis and response becomes critical for extracting optimal value from that data.
AI is beneficial for both real-time and post event processing:
- Post event processing – identifying patterns in data sets and running predictive analytics, e.g. the correlation between traffic & parking usage (see smart parking solutions), air pollution (see air quality monitoring solutions), urban heat island effect with bin sensors and chronic respiratory illnesses within a city centre.
- Real-time processing – responding quickly to conditions and building up knowledge of decisions about those events, e.g. Abibird solution to monitor elderly patients, building up profile based on unique activities in the home to detect and send alert in case of abnormal behaviour.
In fact, it is machine learning (a component of the more generic AI term) that provides the ability to detect patterns in data presented. It learns from these patterns in order to adjust the ways in which it then analyses that data or triggers actions. With machine learning embedded into an IoT environment, you get more ‘connected intelligence’:
- Predictive analytics – ‘What will happen?’
At the most basic level AI enables the power of prediction to forecast and mitigate risky events. This enables organisations to use real-time data to determine when machinery and equipment is likely to break down – so action can be taken to prevent failure and its associated cost via pro-active intervention.
- Prescriptive analytics – ‘What should we do?’
A second level is the power of prescription. Equipping intelligent sensors with logic to drive action means outages or disasters can be avoided. For example, railway track sensors can warn against track failures, or autonomous vehicles can course-correct when the car veers away from the centre of the lane.
- Adaptive/continuous analytics – ‘What are the appropriate actions or decisions? How should the system adapt to the latest changes?’
At an even more advanced level, AI with IoT can deliver an adaptive or autonomous response. This means solutions can incorporate continuous data feeds – where the system learns to take the optimal action without human intervention. So, for example in Agriculture, soil monitoring let’s you determine the nutrients and water level of your soil, so currently farmer can precisely determine the level of fertilisers to be applied and in which areas. But with autonomous tractors, we could easily imagine that all this could be done automatically and precisely and even taking into account the weather forecasts from your own connected weather station.
To give another example that takes into account the 3 parts above, let’s have a look at HVAC monitoring. Predictive: by closely monitoring HVAC systems, you can establish patterns that will determine its over power consumption, malfunctioning or even failure if you wait too long. Prescriptive: the units could be serviced at the optimal time, not too early because it would be near pointless but just before it starts overconsuming, consequently saving on maintenance cost and energy consumption. Adaptive: Imagine, taking into account multiple sources of data such as weather conditions or traffic flow into the building, for even more precise HVAC usage & potential issues with a system that could also clean itself when required. We are not far but not there yet.
Potential Future Uses for AI-Powered IoT Devices
Today’s IoT applications are useful in understanding trends, as they lay out areas in which “traction” is proven and directions where big-company and venture money is already moving. Beyond driverless cars or predictive maintenance of equipments at the tip of the iceberg, there are other potential IoT+AI applications. Hereafter a few useful articles for insight into potential combinations of IoT and AI:
- Harvard Business Review: How Smart Connected Products are Transforming Competition
- The Next Web: 2016 IoT Predictions
- NextGov 2016 IoT Predictions
AI and IOT, complementary technology for value creation
IoT, augmented and enhanced by Machine Learning, is effectively multiplying the impact and benefit to businesses who are adopting these complimentary technologies. In fact, AI can be an integral element for success in today’s IoT-based digital ecosystems.
The reason is simple. Combining IoT with rapidly advancing AI technologies can create ‘smart machines’ that simulate intelligent behaviour to make well-informed decisions with little or no human intervention.
As new technology applications emerge where IoT works hand in hand with AI – the resulting innovations are proving how IoT can create new markets and opportunities, disrupt traditional business models and dramatically change the competitive landscape. Thus, the companies looking to make the most of opportunities to positively impact revenues, safety, resilience and customer experience are being fast on the uptake of this powerful pairing of transformative technologies.
We’re now seeing significant investment in the convergence of IoT and AI. Worth reading the article on the Top 25 IoT Startups To Watch In 2019. Those start-ups are concentrating on how to make IoT a growth catalyst for enterprises by designing in AI integration at the platform level. Also a good read is why are AI and IoT perfect partners for growth?
The result is an acceleration in innovation which can significantly boost productivity for the organisations involved. Little wonder then that the IoT and AI markets are developing rapidly and in tandem.
However, just as AI and IoT can ultimately empower organisations with strong value creation, so the same, strong combination is requiring a more adaptive strategic response from the management of those organisations. If data is the new oil and knowledge is power, most managers are still at the beginning of the learning curve, with few implementing small minor IoT projects before even thinking of adding AI capabilities, for the cases where AI indeed complements IoT.
This leads to teams without the real motivation (or committed resources) to drive an actual result. Businesses must pick a major problem to have a high likelihood of being solved. What mission-critical business information are you dying to know, but can’t currently access and that could be resolved by first, having IoT data and second, refining that everchanging data with AI? Then with that in place, how would you implement it in your business? IoT & AI often involved change of processes that must be embraced by staff as well as customers to be successful. This goes beyond value creation, this is digital transformation at its finest, but this is not for everyone…
One last word, we mainly talked about “Narrow AI”, where AI is used in combination with IoT to perfect one task, either perfecting HVAC monitoring or perfecting soil cropping. But the near future lays in “General AI”, where systems could adjust to solve any tasks, taking info from the web, humans or any IoT sensors talking to each other. Will the be the judgement day of the Terminator?
A Short Glossary of AI and IoT
This article has been written with the professional or executive in mind, rather than the researcher. So we provide hereafter a short broad definitions (and related links) of IoT & AI terms:
Internet of things: Network of physical objects that contain embedded technology to communicate and sense or interact with their internal states or the external environment (Gartner). 2 main categories:
- Consumer IoT: the fitbit style solutions attached or close to humans, to monitor their personal environment, behaviour or health. The devices are usually only linked to a personal App.
- Industrial IoT (IIoT): This is usually referring to industrial or smart cities applications, connecting multiple devices to create efficiencies by better monitoring &/or geolocating equipment. The devices are usually linked to an IoT platform, which is usually integrated to CRM or ERP systems.
Artificial intelligence: an area of computer science that deals with giving machines the ability to seem like they have human intelligence and being able to identify some human senses such as hearing (eg speech recognition and interaction) and sight (eg visual or facial recognition). It is divided into different sub categories:
- Machine learning: a method of data analysis that automates analytical model building. Using algorithms that iteratively learn from data, machine learning allows computers to find hidden insights without being explicitly programmed where to look (SAS), eg. predictive maintenance.
- Deep learning: a branch of machine learning based on a set of algorithms that attempt to model high-level abstractions in data by using multiple processing layers with complex structures, or otherwise composed of multiple non-linear transformations (Wikipedia), eg. IBM Watson debating capabilities
Want to know more about AI, read this article.
Watch recording of the Morgan Stanley Inaugural Australia Summit -June 2019 where Catriona Wallace (CEO @Flamingo AI), Cheesun Choong, (CTO @ Hyper Anna), Scott Helfstein (Global Co-Head of Market Research and Strategy @ Morgan Stanley) and Renald Gallis (VP Marketing & Ecosystem @ Thinxtra) were talking about AI & IoT as the new current disruptors.
Thinxtra is empowering the Internet of Things in Australia, New Zealand, Hong Kong and Macao by operating the world-leading Sigfox LPWA network as well as enabling a full eco-system of IoT solutions and services partners to connect the non-connected, to increase productivity, to accelerate decision making, to improve quality of service and quality of life, and to find more economical solutions to common problems with the ultimate aim to create more efficiencies in a carbon constrained world.
PR Contact: Renald Gallis – VP Marketing +61 404 894 960 email@example.com