After investing in over 60 ICOs in 2017 and making a 20,000% return, our private fund has been almost 100% in cash since January of 2018. It was not difficult for us to spot the peak of the bubble and exit to preserve capital. Finding startups with good potential in blockchain is similar to looking for a needle in a haystack. However, we recently found a needle. I explain below why we have invested in the seed round of http://www.anylog.co/.
Over the last few decades, the world has been through many technological transformations each completely reshaping the world we live in. A major recent transformation is the Internet of Things or IoT. For many organizations, IoT is and will reshape the processes, products and how business is conducted. IoT is about data collected from connected devices and machines and is used to create efficiencies, enhance product designs, develop new business models and improve customer experiences. IoT is comprised of smart devices and machines interacting and communicating with other machines, users, applications, environments and infrastructures. As a result, huge volumes of data are being generated, and that data is being processed into useful actions that can “command and control” things to make our lives much easier, efficient and safer — and to reduce our impact on the environment.
Examples of IoT devices range from sensors on traffic lights to complex industrial machinery. These sensors are everywhere — for example supersized Airbus A380–1000 airliner is equipped with 10,000 sensors in each wing. The number of IoT devices worldwide is predicted to grow to an astounding 31 billion by 2020 which will be 4 times greater than an estimated 7.3 billion smartphones, tablets and PCs. With a growing number of IoT devices generating exponentially more data the question becomes how will all this data get stored and processed?
Today, most of the IoT data is stored at data centers in the cloud. However, most experts agree that these data centers are not well suited to manage the scale IoT data. A cloud is a centralized data center that offers processing and storage capacity to service applications and data. But as each data center is managed by a company with finite resources there are constraints on the amount of compute resources that a data center can make available. Some companies like GE attempted to address the problem by building very large data centers optimized to service IoT data. While this approach makes sense for some IoT use cases, this approach fails to deliver a generic low cost alternative to manage IoT data. In many cases, the cost of bringing the data to these data centers exceeded the economic value that was extracted from the data and therefore, only a small fraction of the IoT data is managed by these data centers.
Another problem with a centralized cloud approach is its ability to support applications requiring real time interaction with data. For example, in a power plant a pump needs to be turn on if the temperature exceeds some threshold. However, in order to manage data in the cloud, the data needs to be transferred to the cloud, re-structured and loaded to databases and only then it could be queried. These processes take too much time for many use cases. For these reasons, companies like Cisco have suggested an alternative approach where data is processed next to where its being originated. This approach is often called edge computing as the data is managed at the edge of the network near the devices that generate the data. But this approach is not simple — companies now realize that with the edge approach, they need to manage enormous number of small data centers spread around the world rather than a single data center at the cloud.
Regardless if the data is managed at the cloud or at the edge, the management of the IoT data requires enormous compute resources with significant management overhead and cost when offered by the companies that manage data centers. By contrast, individual computers are everywhere. People and corporations maintain enormous numbers of computers which are not being fully utilized. What if we could somehow leverage these computers to store and process IoT data? This is the idea behind the AnyLog platform. AnyLog is a network that is managing IoT data by distributing it to many individual computers that are part of the peer-to-peer AnyLog network. The network creates incentives for people and corporations to contribute compute resources by providing payments (in the form of tokens) when their resources are used. With this approach, IoT data is managed by many connected computers which operate, using the AnyLog software, as a single supercomputer. The blockchain is a key component in how the AnyLog network works. The Blockchain is a distributed ledger containing a chain of blocks with each block in the chain containing information that can be shared with any member of the network. The AnyLog technology uses this property to bind all the compute resources of the individual computers in the network such that these computers process the data as a single supercomputer to efficiently manage enormous amounts of IoT data.
With this approach, when devices generate data, the opportunity to store this data is offered to network participants in a bidding process where participants interested in monetizing their resources offer to service the data. For example, a device communicates with the network and requests storage and query processing for 3MB of data for 2 months. Participants of the network reply offering their prices and terms of the service. This approach creates a marketplace where service providers are incentivized (by token rewards) to join the network and compete to service the data. A device selects the winning service providers and distributes the data and processing tasks to them. The end result is that the entire IoT data is distributed to individual computers around the globe and users, using the AnyLog interfaces, are able to efficiently query the data. The fact that the platform incentivises data owners to push their data to the network makes the network an integrator of IoT data from many different sources. In the same way that Google provides an unified way to query website data, the AnyLog platform offers a unified interface to IoT data hosted on its network. For example, with this approach a smart city can issue a query to find how many cars entered the city in the last 2 hours and a power plant can issue a query to determine how much heat was generated by a turbine in the last 2 days.
The AnyLog platform provides another important innovation. With a typical cloud solution the data is first brought to the cloud, then inserted into a database through a process called ETL (Extract, Transform, Load). Once the data is in a database, it is possible to run queries on it. However, creating and managing these databases is very complicated and expensive. Using AnyLog, companies will manage data without the need to build and support these backend databases. Companies can just publish their data to the network, and immediately issue queries against their data. This results in significant benefit from removing a need for backend databases with their initial and ongoing management costs.
AnyLog’s application of blockchain technology to address IoT data management challenges is an example of the latest technological transformation of blockchain technology. AnyLog’s approach of creating a network that overcomes the main challenges in handling the enormous volume of IoT data has the potential to change the way IoT data is managed.