How to turn data from embedded systems into a marketable service
Increasingly, product manufacturers are using the data they gather from embedded systems to create services, such as data visualisation tools or predictive maintenance that are useful to their end customers – consumers or other organisations.
By viewing their data as an opportunity instead of a risk to be managed, many companies have successfully commercialised this asset. The development of the Internet of Things (IoT) and the Cloud have enabled this trend and are giving rise to the product-as-a-service (PaaS) business model, whereby customers pay for outcomes rather than products. The manufacturer still owns the product, while the customer leases it, along with added services derived from the data contained within the devices.
But what will product owners need to consider in order to make the leap into the service sector via data integration? And how can companies make the most of the data they own?
Connectivity is key
Bogdan Racotea, European Head of Software Engineering at Nippon Seiki, which designs and manufactures automotive driver information displays and sensors, spoke at an SThree webinar on the PaaS business model. He was clear that connectivity is critical to success: “It will define the fate of the product.”
Companies will need to consider what will power the huge number of connected devices and sensors coming online as the Internet of Things grows rapidly. According to Gartner, the IoT will comprise 25 billion connected ‘things’ by 2021, producing an immense volume of data.
When wireless sensors are used over wide areas and in remote locations – for example agriculture or smart grids – concerns over battery life and maintenance costs are high. Such concerns have led to the emergence of Low Power Wide Area (LPWA) technology, which is designed to allow long-range communications at a low bit rate.
This includes Narrowband Internet of Things (NB-IoT) communications, which enables thousands of remote sensors to operate using only a small amount of bandwidth. Another long-range standard is LoRaWAN, a low-power solution for IoT that is developer-friendly and low-cost.
Also speaking during the PaaS webinar, Alex Lennon, Founder and Director of Dynamic Devices, which provides a range of embedded integration services, said: “If we need to do firmware updates to what sensors are doing, that’s a real difficulty because we have very, very little bandwidth available to us to perform those things. We need to be coming up with some clever strategies.”
Design for security
Embedded security plays a critical role in connectivity. This is something that companies looking to commercialise data need to consider carefully upfront, as the IoT poses fresh security risks at a time when high-profile cyber breaches and state-sponsored attacks are hitting the headlines.
The costs for a business can be steep: a 2020 report from IBM found the cost of a single breach is nearly $4 million, and Racotea says it can “destroy” reputations. Suppliers can also share the blame when things go wrong.
Most of his team’s overheads arise from trying to understand how to respond to the individual security requirements of customers. “Each one of them has a specific take on how to implement the cyber security part,” he said. “And that makes our lives quite difficult.”
At the same time, Lennon senses a “tipping point”, where security is not viewed merely as a cost to be set aside but as a core part of the business model. However, “it’s extraordinarily difficult to put secure systems together” and there is no silver bullet solution.
If you would like to discuss the different skill sets needed for this kind of transitional project
He suggests becoming familiar with government standards such as Secure by Design, saying “we desperately needed some direction in terms of how we secure the products that we bring to market”.
There are some simple steps to take, according to Lennon, for example ensuring you do not use default passwords, planning for the obsolescence of your hardware and an exploit against your devices or systems.
Lennon also recommended engaging with best-in-class security consultants from the design phase of a project rather than as an after-thought. It’s difficult to reverse-engineer a solution, and security is a specialist area.
The bigger picture
Alongside security expertise, Racotea adds that one missing element is project management to bring all the different types of expertise together into a whole that’s greater than the sum of its parts. “Often, the software engineer plays the role of a lone wolf, and goes around and fixes everything.”
Another panellist, David Morrison, Engineering Group Director at Cadence Design Systems, a leading provider of electronic design automation (EDA) tools, stressed the importance of marketing skills, so as to educate customers on how they would benefit from product-as-a-service models. “We need to be able to get that message out.”
Data visualisation can create real business value, for instance, but Lennon says customers “don’t want to be overloaded with massive amounts of information. They just want to know what they want to know. So it’s critical to focus on understanding that customer need and simplifying down the data visualisation”.
With more and more devices becoming connected to the internet, he says the goal should be to draw wisdom and present insights from the deluge of data, prioritising data quality over quantity.
Morrison points out that it’s critical to ensure that product-as-a-service models will benefit customers rather than being used only as a method of monetization by companies. “I think with any product there has to be a market need. People need to need to want it... It needs to solve a problem.”
If you are transitioning to a product-as-a-service model and would like to discuss what skillsets are needed, contact Computer Futures today!