the first commercial-grade system for curating sensor data on a global scale. Terbine is designed to act as the trusted sensor data exchange for and between sectors, including both commercial and governmental entities. Terbine is a highly sophisticated cloud-based system that makes possible fast, frictionless distribution of IoT data, and covers the core issues involved with monetization, regulation and licensing. And Terbine can be white-labeled by large companies to enable private data marketplaces.
The company is still in beta and testing its offering with publicly available data sets before it rolls out any commercial services. David Knight lists some of the key challenges his company faces and which will be common to other IoT data exchanges:
- sourcing unique data sets;
- cleaning the data;
- applying metadata to it in a uniform way;
- persuading third parties to share their data on the exchange;
- building in safeguards to restrict what can be done with the data by Terbine users;
He notes that while it is difficult to persuade many companies of the benefits of sharing their data, or “crown jewels” as he puts it, most of them seem fairly happy with the idea of using Blockchain technology to validate transactions on Terbine. Overall, it seems like the company are really ironing out a range of very complex issues before going live. Vibrant, commercially successful data exchanges like Terbine will be essential for the IoT to have the impact many are predicting. Whether the market will naturally go to several mega corporations such as Amazon and eBay in the consumer ecommerce space or if it will support a number of smaller exchanges remains to be seen. Network effects will play an important part, buyers will follow sellers and vice versa, but it is possible to imagine a network of exchanges using common APIs and metadata to provide seamless services to end-users.
Interesting article in this morning’s FT (requires subscription) about the amount of energy used to power data centers. According to the author, Izabella Kaminska:
The information and communications technology (ICT) sector absorbs up to 10 per cent of the power resources of the UK’s National Grid. Of that, one-third of the power goes to supporting the network, one-third towards supporting our devices and the other third goes towards supporting data centers.
The solution, according to Kaminska is for data which is unstructured, not classified and not revisited after uploading to be deleted in the same way our brains forget things which are no longer important to us. The main offenders, she argues, are all the funny cat videos uploaded to YouTube and other sharing sites. Apparently, a few years ago it was pornography that was clogging up the internet but now it is more innocent content.
As the Internet of Things (IoT) starts to permeate our lives we can expect data storage to be an even greater issue. While much of this data will be structured and not have the storage requirements of large video files, it will be the sheer quantity of data flowing through the network that will challenge data center operators. However, I feel this is an issue where technology will solve the problem it has created. More efficient storage and access methods, greater energy efficiency of the servers and renewable energy sources will make incremental but important contributions. It is not clear that a system of better classified cat videos is practical at either the production or consumption ends of the chain.