A report, made for the U.S. National Institute of Standards (NIST),
Similar results to what Eclipse IoT finds in their survey.
Foundational gaps that hinder the scaling of IoT:
| Interoperability | The lack of interoperability hinders different devices and systems from integrating, communicating and sharing information with each other. Barriers to achieving interoperability include limited focus of standards initiatives, resistance to open and industry consensus standards, regional standards and standards implementation errors and deviations. |
| Cybersecurity | |
| Privacy | |
| Connectivity | Connectivity challenges are multi-dimensional in nature and challenging to solve due to various factors, including the need for substantial infrastructure investment, the lack of a “one size fits all” approach along with issues in market economics, funding and incentives, “last acre” coverage and spectrum. |
| Data management challenges | As IoT scales, so does data management complexity. The IoT data collected comes in a variety of types, formats and sizes. Some data are time-sensitive and must be processed immediately while others are stored for future actions. Data may be required to comply with industrial, state and national regulations. Robust data management is foundational for artificial intelligence systems. Data management challenges are complicated by exponential growth in data volume and velocity, privacy considerations, cybersecurity factors, interoperability concerns and regulatory compliance requirements. |
| IoT data ecosystem | The future IoT data ecosystem is envisioned to be a highly interconnected network where data generated by IoT devices and systems is seamlessly shared, monetized and utilized across various sectors. There are, however, a number of technical challenges to building such an IoT data ecosystem. These include data quality, interoperability, privacy and security, data sovereignty, scaling and standards for data management. |
| Communications and network infrastructure | Industry analysts estimate that there will be 55.9 billion IoT devices generating 79.4 zettabytes (ZB) of data by 2025. Current communication networks and architectures are not designed to manage the needs of IoT at this scale. New processes and technologies for configuring, managing, operating and maintaining the hyperconnected network will be necessary. Representative areas of infrastructure innovation are needed to support real-time autonomy and complex IoT applications, be fault tolerant and resilient and defend and heal against threats. |
Good quotes:
P. 28: There are millions of legacy and OT systems in use today, from manufacturing machinery to Programmable Logic Controllers (PLCs) and SCADA (Supervisory Control and Data Acquisition) systems. While some of these legacy and OT systems may offer data collection and control capabilities, they were not designed to connect to and communicate across the Internet.
It is neither feasible nor practical to replace all these legacy and OT systems with new connected “smart systems”. Some of the legacy systems must be retrofitted with IoT technologies to enable them to connect, communicate and be interoperable with existing systems and modern smart systems. An ecosystem of solutions providers who build “bridging” solutions is required.
p. 29 (Edge computing / Problem with device-to-cloud): In a traditional IoT architecture, data is routed from the device to a remote cloud data center for processing and storage. Not all data collected needs to be or should be sent to the cloud for processing. In mission critical applications or in those where connectivity is intermittent or unreliable, processing is performed at the gateway or by the device itself. A 2022 survey of 910 IoT developers’, conducted by the Eclipse IoT Foundation, found that the top computing workloads performed at the edge were artificial intelligence (38% of respondents), control logic (34%), data exchange across multiple nodes (22%) and data analytics (20%).
P. 30: While this approach leads to lower development and ownership costs, continuous innovation, improved code functionality, performance and resilience and mitigation of vendor lock-in concerns associated with proprietary software, there are implications for long-term maintenance and cybersecurity updates.
P. 35: Interoperability
- In transportation and logistics, the lack of universal standards for freight systems hampers data exchange, causing supply chain delays and increased costs.
- Smart cities, filled with IoT devices and systems owned by different entities, struggle with interoperability, locking cities into specific vendor solutions and preventing efficient data exchange between systems.
- Healthcare also suffers from fragmented, incompatible systems that make it difficult for medical devices to communicate, slowing down care and reducing operational efficiency.
- In renewable energy, interoperability is crucial for grid reliability, but inconsistent standards and policies hinder the integration of distributed energy resources (DERs) and energy management systems into the grid.
Ultimately, across these sectors, the lack of universally adopted standards and reliance on proprietary systems prevents seamless communication and data sharing, leading to inefficiencies, higher costs and reduced innovation. Without a concerted effort to adopt open standards and improve interoperability, these industries will continue to face challenges in maximizing the potential of their technologies.
As IoT scales, so does data management complexity.
Robust data management capabilities simplify these challenges and help unlock the value of IoT by enabling massive amounts of data to be collected, processed, stored, discovered, queried and analyzed.
P. 37: However, data management faces a variety of challenges:
- In the construction industry, the integration of IoT with Building Information Modeling (BIM) systems is challenged by siloed and fragmented data from various contractors, their reluctance to share data and a general lack of trust in a fragmented value chain.
- In transportation and logistics, managing the vast data generated by IoT systems in the global supply chain is essential for smooth operations. With data flowing through multiple touchpoints such as manufacturing, transportation and warehousing, the challenge for supply chain and technology managers lies in handling decentralized and diverse data formats, while adhering to various regulatory standards. The exponentially growing volume of data requires businesses to have a scalable infrastructure capable of processing and storing information effectively.
P. 38 Industrial: In manufacturing, factories are faced with a challenging wireless environment, where machinery, metallic surfaces and concrete structures interfere with signal propagation and create latency. Many manufacturing operations require real-time data monitoring, making on-device or local gateway processing essential. Additionally, many factories lack the necessary network infrastructure to support connected operations, further hindering their deployment of edge computing solutions.
The high cost of implementing IoT presents a major barrier to their adoption across industries like manufacturing and retail. A large manufacturer may operate multiple factories containing thousands to hundreds of thousands of machines and production equipment of varying types. The cost to fit all production equipment in these factories with various devices can be cost prohibitive. The total IoT investment required is more than just devices, as hardware is usually around 30% of the initial implementation cost. Additionally, the total cost of implementing IoT includes not just the devices, but also infrastructure upgrades, professional services and recurring costs for cloud services. Small and medium-sized manufacturers face even greater challenges due to limited capital and outdated infrastructure.
P 39: For smart cities, edge computing is crucial for managing the growing number of IoT applications that demand real-time data analysis and processing at the device, gateway, or local server level. However, technical advances are needed in areas like scalable architectures, energy efficiency, interoperability and cybersecurity to fully realize the potential of smart city technologies. Without these improvements, cities will struggle to support the vast amount of data generated by IoT devices, limiting the effectiveness and expansion of smart urban environments.
P. 77: Interoperability: 6.1.1
P 91: Connectivity – spectrum 6.1.4.3.
6.2.1. Intelligence: Data management
The problem with Interoperability has been solved by the royalty-free layer-3 protocol Z-Mesh: https://z-mesh.org/
For IoT to succeed, it must be:
- Scalable (async to support offline devices): have a look at why Apache Kafka is a success.
- Interoperable (Network Effect): Any IoT-event (eg. temperature reading) must be accessible to any other client
- Support for both Constrained and TCP/IP devices
- Royalty-free: Not patented by anyone. It must be an equal market in which any vendor can contribute sensor or app or networking equipment – look at the success of the Internet
- Direct Devices-to-Device communication: To support Edge and low-latency AI applications, sensor-to-cloud technologies are not good enough.
Download the report here: https://strategyofthings.io/nist