Announcing SORACOM MOCAROS, a breakthrough service for IoT data inversion

Soracom Mocaros Icon

While many services are now available to allow storage of IoT data, all of them are limited to accurately storing data as it was originally input or recorded. Since the world as it is can be a little disappointing, this limitation has proven to be a significant barrier in overall adoption of IoT systems, which of course promise a better, more connected world.

It is possible for users to invert data sets by hand, but the process is somewhat tedious, requiring a download and validation of stored data, a potentially lengthy processing cycle, and a fresh upload of the handcrafted data. In cases involving small data sets, this might be merely irritating. In cases involving large data sets, it can mean a project is shelved before it even gets through testing. Imagine, for example, a “quantified self” system that lets users literally turn their lives around by overwriting their disappointing real-life metrics with a completely inverted set of fabulous metrics.

Today, Soracom is happy to announce MOCAROS, a breakthrough solution for automated inversion of IoT data. MOCAROS uses a proprietary space-time machine learning technology (Artificial Preferred Reversible Intelligent Learning or “APRIL”) to identify the core characteristics of data as it is entered, and not only simply reverse but invert the data set on the fly while maintaining its fundamental structure.

This allows users to perform basic functions like changing numbers from positive to negative; intermediate transformations, like inverting complex image data to turn, for example, water into wine or oranges into apples (allowing apples to apples data set comparisons); or advanced functions like turning sad results into happy outcomes. This last case is a good example of APRIL’s capabilities – simply reversing the data without maintaining its structure would reify an already poor result, transforming it from “sad result” to “Das result.”

While MOCAROS is clearly quite powerful, it takes advantage of existing Soracom capabilities to ensure ease of use. Users of Soracom Inventory can simply create a new device definition, allowing direct data entry into MOCAROS via Soracom Air or any other internet connection. Similarly, data inverted by MOCAROS is available to be transferred to Soracom Harvest for storage and visualization.

For complete documentation, you can check out the article Send data to Soracom MOCAROS using device ID and device secret on the Soracom Dev Site. For this blog, we’ll look at a few simple use cases:

Image confirmation via data mirroring

If you use MOCAROS to reverse image data on the fly, you can still validate the original image data by viewing the Soracom management console in a mirror.

Palindrome confirmation

You can easily scan incoming data for “data palindromes” that remain the same in both original and inverted states. If the data uploaded from an IoT device is always a palindrome, that IoT device is likely to either have been blessed by a deity or (more likely) hacked. Palindrome confirmation makes it easy to detect potentially compromised devices. And of course “SORACOM MOCAROS” is also a palindrome.

Turn losses into gains and defeat into victory

In many cases, the world would be better off if things had gone a different way. But with MOCAROS, you don’t have to worry. Economic downturn or stock market drop? Feed global data into MOCAROS for an instant recovery or bull market. Don’t like where the country is headed? MOCAROS is here to take away the election-night blues (or the reds, depending on your preference).

 

FAQ

When is Soracom Mocaros available?
Soracom Mocaros is available in public preview beginning April 1, 2019.

How much does it cost to use Soracom Mocaros?
There is no charge for Mocaros. However, there is a cost associated with Harvest and Inventory at higher levels of use so please stick to the free tier.

What are the benefits of data inversion?
Although the benefits of data inversion will differ depending on the customer, we believe that the risk of not reversing is greater.

Is simple image flipping supported?
There is no current support, but this may be supported in the future.

Help! When we ingested inverted data from MOCAROS, we got a large number of alerts from our back-end system.
Excellent! That confirms everything’s working as designed. if you do not receive alerts, you should be concerned that there might be a problem with your back-end system.

Are you planning to patent this technology?
Mmm… probably not.

If you want to use MOCAROS in a large amount, is it necessary to apply for use volume reduction in advance?
We are very interested in this use case, so please contact us.

 

Endorsements

Fusic Co., Ltd.

We offer the “mockmock” virtual device creation service for IoT systems. The mockmock virtual device can inject arbitrary data into various Soracom services, making it easier than ever to debug, test and load IoT systems without having physical devices.

Combining SORACOM and mockmock can make it fun to develop high quality IoT systems. However, since we first released the mockmock service, many customers have asked for a way to somehow reverse the data and send it back. Normally we would hesitate to handle this kind of advanced space-time technology, but based on the evidence we saw in documentaries like Super Dimension Fortress Macross we felt that it might be possible to solve this with songs and deliver the songs to the customers.

With today’s release of SORACOM MOCAROS, it has become possible to meet customer requests without implementing advanced space-time functions on our side. We will continue to contribute to the acceleration of IoT system development, with a focus on MOCAROS and mockmock.

Shintaro Sugimoto
Fusic Co., Ltd. Manager

 

RealtimeSystems, Inc.

RealtimeSystems provides location information solutions that are easy to deploy to small and medium-sized businesses using high-precision, long-running GPS trackers.

Our solutions leverage the global coverage of SORACOM Air to provide a variety of location-based solutions including protecting the elderly, preventing theft, and managing bus operations. With SORACOM MOCAROS, it is possible to invert outbound GPS trajectory data  and automatically generate return-journey destination based on travel in the opposite lane. Such inversion simulation enables two-way behavioral analysis.

Overseas traffic rules that differ from those in Japan, such as right lane driving in the United States, can be automatically identified by SORACOM MOCAROS to enable left-right reversal of the one-way trajectory, and such behavior analysis functions can be provided quickly.

RealtimeSystems believes that continuing to provide reliable, global GPS solutions will be key to creating an enterprise-class Big Data use case for location data.

Kanami Saeki
CEO, RealtimeSystems INC.

Brains Technology, Inc.

Brains Technology provides “Impulse”, an advanced anomaly detection / prediction detection solution that utilizes machine learning in manufacturing, construction, network monitoring, etc.

Directly linking “Impulse” via SORACOM Funnel makes it easy to handle IoT sensor data installed in a wide variety of environments, reduces data costs, and ensures security. This makes it possible to implement a variety of predictive functions such as detecting equipment issues or product defects, diagnosing performance degradation over time, or performing other factor analyses.

Data inversion can be used to extract key feature quantities from time-series data, but requires an automated data processing mechanism for efficient performance. The release of SORACOM MOCAROS makes it possible to process data in reverse without any custom development. In addition, we can now use space-time inversion to pre-identify anomalous conditions before they occur, using a space-time transition learning model based on APRIL (Artificial Preferred Reversible Intelligent Learning).

With “Impulse” and “SORACOM MOCAROS”, we will contribute to advanced IoT data utilization.

Toshiaki Enami
Director of Brains Technology, Inc.