Authors: Brendan Rowan, Tanya Suarez | BluSpecs
The evolution of the IoT into the Next-generation IoT will be marked by technological advances that include contextually aware and distributed intelligence, federated services and microservices, integration of DLTs and smart contracts, and tactile and AR interfaces. It signals the meeting of cloud and edge infrastructures and a human-cloud computing continuum, underpinned by the functionalities afforded by 5G networks.
Under the NGIoT Initiative there are several R&D projects developing the architectures, microservices and interfaces, demonstrating the value of the Cloud-Edge-IoT (CEI) paradigm and across use cases, ranging from health and farming to manufacturing and ports. During IoT Week 2022, we hosted a panel session which sought to identify and address the opportunities and challenges surrounding the supporting the uptake of these technologies and the unlocking of genuine and serious investment in IoT technologies.
The road to the next-gen IoT
It could be argued that the IoT ecosystem is already defined by fragmentation, with overly diverse definitions, architectures technological approaches and ambitions. In this context, the approach of cloud-edge deployments will not be for the uninitiated or for the faint of heart.
Opening up cloud-edge opportunities to a wider community will require common standards interoperability, trust and privacy and possibly even an exploration of open source. Over the next five years, collaboration will be key, with distributed or swarm computing capabilities and data flows designed to reflect how markets and communities function.
The regulatory framework, which includes the proposed Data, AI, Digital Markets, and Chips Acts, is evolving. It will guide the development of technologies and their practical applications in line with European values. This raft of policy instruments is set to introduce new roles and bodies that will affect how the market will develop, and how fundamental and non-fundamental rights are exercised, triggering alterations to systems as they are developed today.
Clarity, relevance and relative ease of compliance will be key to ensuring that the competitiveness of European Cloud-Edge firms can be sustained in Europe and other world regions.
Challenges in stimulating the edge investment
The deployment of edge computing and increased intelligence of devices is set to pick up pace as the benefits for businesses, consumers and citizens become clearer. There are some complexities, however, that could hinder this process.
1. Access to edge infrastructure
For one, many stakeholders still do not have reasonable access to fit-for-purpose edge infrastructure; startups, SMEs and researchers are struggling to affordably build and deploy on the edge. This is further compounded by skills shortages; the talent of the future is required to be technically competent, business-aware and have at least a basic understanding of the legal context. This combination of skills is currently addressed at fourth-level education but must be brought into undergraduate technical and non-technical courses.
2. Clarifying data ownership
Clarify data ownership, and corresponding rights and responsibilities in more complex data environments, is a priority. In the context of vehicle-to-everything (V2X) communication, the advanced processing of data within an electric vehicle and connected car raises the question of the ultimate owner of the data generated and models developed: the driver, the owner, the ECU manufacturer, the internal data platform provider, the cloud service processor, the OEM?
3. Domain-specific hardware and software
Some parallels can be drawn with the Aerospace industry who has historically led IoT and edge processing through the integration of fly-by-wire systems and power-by-hour business models. In fact, it is reasonable to assume that the high-level architecture required for the connected and intelligent car will be similar to that of an Airbus aircraft. In contrast, in the view of some panellists, processors are likely to remain sector-specific for some time, as will their abstraction.
“The Aerospace industry is extremely unlikely to accept standards that have been developed for the Automotive industry”, Florin-Calin Paun, Chief Science Officer, RAFAUT Group.
4. Trust in the technology
Ethics and trust are at the core of driving adoption and development of technical solutions, if we are to progress to the acceptance of autonomous intelligence, where the device takes decisions without human intervention, we must be able to trust the models that lead to that decision being made, and achieve reliability at 99.9999%. In this respect, the use of the term IoT may in itself be a barrier, with the term Internet reducing trust among the general public through association with social media and current societal concerns with privacy and use of personal and business-critical data.
“The law is the floor, not the ceiling when it comes to personal data”, Daragh O’Brien, Managing Director, Castle Bridge.
Our shared culture is important in achieving trust but there are nuances in the approach. The regulations (e.g. GDPR) provide the floor, not the ceiling when it comes to establishing rights and serve to express the minimum translation of cultural differences within Europe. The focus for all tech developers should not be on liability but on the visualisation and consideration of the potential impact of the solution on people and society as the end from which to work backwards.
5. Continued evangelisation and education
At the moment, the IoT products and solutions still require evangelisation and education, as most IT and OT spending does not easily align with IoT spending in more traditional organisations. Vendors are required to provide end-to-end solutions that combine the provision of connectivity, hardware and applications. In fact, ‘use cases’ are still the core of the discussion and remain fairly simple in comparison to the complex advanced contextual IoT business value that will be created through the NGIoT.
Startups provide the vehicles for developing and driving these use cases, but they, like the Beatles, are often required to go to the US to scale. In the view of one panellist, the impact is significant as founders and leaders shift focus to comply with US VC demands.
Another panellist drew attention to the lessons to be learned from the telecoms market in the late 80s, and early 90s which resulted in a race to the bottom and a stifling of innovation.
“Sensor value is 24%, the connectivity is 9%, the rest is platform. The connectivity will end up being free but where does that leave the infrastructure providers”, Paul Delaney, Head of Sales, Cellnex Ireland.
Reaching European Cloud-Edge-IoT goals
Firstly, we need to stop talking to businesses about edge computing, IoT, the cloud, etc. Generic conversations and descriptions do not communicate the value of adopting the NG-IoT to business leaders across different domains: whether an autonomous vehicle turns left or right based on local processing or cloud is not of relevance to the driver. The driver just needs to know that the vehicle will respond to driver’s need to get from A to B. These needs must inform the development of platforms and systems now.
We can learn from the past. The first fly-by-wire aircraft crashed because of the information overload on pilots, essentially due a lack of understanding of the human-computer interface. Now over 500,000 parameters in the modern jet are combined and reduced to just 100 meta parameters that reach the pilot.
It is estimated that up to 30% of a company’s turnover is lost annually due to poor quality data, through poor decision making and inefficiencies. The IoT turns on the hosepipe for data, and while moving the processing closer to source will provide more functionality, standards are key to ensuring data quality. We must also develop the tools the tools that allow non-expert users to extract value and generate opportunities.
Hard policy can be employed to incentivise beneficial business models and disincentivise some of the more harmful models that have proliferated over past decades. Softer policy approaches can be leveraged to trigger investment, accepting risk and investing heavily for clear commercial outcomes. As ever, the balance between regulation and stimulating innovation must be pursued.
“What is currently under constant discussion within the EC and with the coordination projects is where is the crossing point at which the value of the technology is sufficiently proven so that private industry will take a large risk and invest heavily”, Rolf Riemenschneider, Unit E4, DG CNECT, European Commission.
Finally, let us avoid the magpies (who hoard shiny projects), pigeons (noisy and messy) and swans (beautiful, but with much effort under the surface required to progress relatively slowly) and seek the Golden Eagles (projects that soar and gain momentum after launch) It is these Golden Eagles that will unlock the private investment, that will lead to the success of the cloud-edge-IoT paradigm.
