Many obstacles are slowing down the adoption of the IoT.
First, the market for IoT devices and platforms is fragmented, with many standards and many vendors. There is ongoing uncertainty about the technology, the vendors and the solutions offered.
Second, there are concerns about interoperability, as the solutions implemented often tend to create new data silos.
Data in the cloud is often stored securely, but cloud-based security implementations cannot protect your data against devices with compromised integrity, nor against data tampering at the source.
Finally, the centralized architecture of most IoT solutions means that there is serious potential lack of resilience. When all transactions are processed in the cloud, unavailability of cloud resources can freeze your business operations.
Blockchain is an emerging technology that could help with system resiliency.
The basic concept of blockchain is quite simple: it is a type of distributed database that is referred to as distributed general ledger , and it maintains a continuously growing list of ordered records. (See a blockchain video demo: https://anders.com/blockchain/.)
The press usually ties the term “blockchain” to transactions, smart contracts or cryptocurrencies, but the technology itself has a much greater potential. This is why we need to dissociate blockchain from specific implementations such as Bitcoin and Ethereum. For example, Walmart and Ford are using blockchain to improve supply management, with no cryptocurrency involved. In fact, the convergence of blockchain and the IoT is on the agenda for many companies. And there are existing implementations, solutions and initiatives in several areas outside of the IoT and financial services.
According to IBM, the three benefits of blockchain for the IoT are building trust, cost reduction and the acceleration of transactions:
- Building trust between the parties and devices with blockchain cryptography and reducing the risk of collusion and tampering
- Reducing cost by removing the overhead associated with middlemen and intermediaries
- Accelerating transactions by reducing the settlement time from days to nearly instantaneous
How would a blockchain-based system accomplish all of this? IBM’s point of view is that all devices in the blockchain should have the resources to run the blockchain software. With every element in an IoT system able to process blockchain data, suddenly blockchain becomes the solution to every problem! Well, not entirely.
A key element of distributed ledgers is that they are open; they are not usually ‘owned’ by any one entity. Any computer connected to a distributed ledger is called a ‘node’. Most of the nodes are lightweight (or at least lightweight relative to the capacities of cloud servers), and so they don’t hold the full ledger. Each ‘block’ within the ledger has a maximum size of 1 MB. A small desktop computer can easily hold a full copy of the ledger, but this is not the case with the majority of IoT devices. Any blockchain system needs at least a few ‘full nodes’ containing the complete ledger.
Another issue is that you need the proper security credentials to view a transaction. So, IoT security issues are still present when using this technology. Device commissioning and secure key management are still ongoing issues with IoT devices, and not solved by blockchain.
Assuming that these limitations are addressed, blockchain would add a compelling commercial dimension to the IoT. A block contains the transaction but can also contain the contract. So, an IoT device could buy or sell data from/to another device or system. This could produce a transactional system less prone to the problems of resiliency.
Yet despite these benefits, the blockchain model is not without its shortcomings. These include:
- Scalability. Blockchain does not yet scale very well. This might lead back to centralization, defeating the purpose of the distributed ledger.
- Processing power . Small devices do not have the power required to perform encryption for all the objects involved in a blockchain-based ecosystem. The forecasted billions of IoT devices will be produced in very large volumes and at very low cost, and the majority of these devices will not be capable of running the required encryption algorithms at the desired speed.
- Storage . Even if blockchain eliminates the need for a central server to store transactions and device IDs, the ledger has to be stored on the nodes themselves. The ledger will increase in size as time passes. This is beyond the capabilities of a wide range of smart devices such as sensors, which have very low storage capacity (either internal flash memory or external NOR or NAND flash).
- Lack of expertise. Few people understand how blockchain technology works. Mixing blockchain and IoT technologies adds great complexities to a system.
- Interoperability issues. It is well known that the value of the IoT rises when you can combine data sources. We already lack data model standards for many vertical markets. Adding blockchain to the mix will only make this issue more difficult, not to mention the legal and compliance issues that such transaction management will create.
The blending of blockchain with the billions of IoT devices is not for the immediate future. Blockchain processing tasks are computationally difficult and time-consuming, and IoT devices are still relatively underpowered, lacking the processing power to directly participate in a blockchain. This limitation is for a good reason: the heavy computational load helps protect integrity. As Salil Kanhere, an associate professor and researcher at the University of New South Wales, puts it: “Standard IoT devices can’t do this kind of heavy computational work, just like you can’t mine bitcoins on a standard laptop anymore.” So initially, this type of application will be seen on high-end gateways first.
The industry press is now reporting that AI and blockchain are accelerating the growth of the IoT, but I believe it is the other way around. The IoT is now making it possible to provide a framework where blockchain can be useful. For the vast majority of edge nodes, blockchain functionality will initially need to be implemented in a gateway, a hardware device acting as the point of transaction. This adds an additional layer of complexity, as this means that the edge node responsible for the transaction would not be the device that records the transaction. The edge node would be a peripheral to the gateway.
Even if the combination of IoT and blockchain is not ready for today, it’s time to start looking at that convergence. Blockchain services for the IoT could trigger the open exchange of data among users and the breakdown of silos that limit access to useful data. We are seeing the beginning of the application of these new technologies. It is only a matter of time before the proliferation of low-cost processors with adequate resources will feed the continued improvements in IoT systems required for blockchain deployment.
A good example is Filament, a startup proposing a new chip enabling connected devices and equipment to complete transactions on their own and create contracts recorded on a blockchain based on rules with procedures set by their operator.
An open-source version for blockchain technologies is also provided by Hyperledger, a global collaboration hosted by The Linux Foundation.
— Christian Legare is the Director of IoT OS Platforms at Silicon Labs. He joined Silicon Labs through the company’s acquisition of Micrium Inc., a leading provider of embedded software, including the µC/OS real time kernel, where he was the CTO since 2002.
From the article "Data in the cloud is often stored securely, but cloud-based security implementations cannot protect your data against devices with compromised integrity, nor against data tampering at the source." The block chain takes the data from the device. If the data from the device is flawed, wouldn't the block chain be no more effective than a cloud-based implementation?