SAP Startup Focus member AppOrchid brings computer cognition to IoT
Krishna Kumar, founder of AppOrchid joined us at this year’s DEMO Fall conference. Read below to hear from Kumar more about AppOrchid solution, and how it will impact the Internet of Things.
Amidst heightened awareness of climate change and geopolitical disputes associated with the fossil fuel supply chain, there is a growing mandate to implement renewable solutions to counter these threats. The national grid of Israel was tasked with directive to deploy solar stations and increase the number of renewables across its grid. However, solar production is not predictable and can cause significant intermittencies leading to forecast issues and challenges in pricing and operational management. Since there is currently not a systematic approach to capturing metrics, the system planning that ensues is inaccurate, rigid and capital intensive.
The approach used for portfolio planning today is very elementary, with little understanding of the variables at play. As dependency on renewable climbs, the legacy tools utilized are unable to incorporate the empirical assessments of subject matter experts, simulated weather conditions, and variable macroeconomic and geopolitical attributes. The inability to perform system planning within a certain degree of accuracy leads to several millions of dollars in lost opportunity and reliability.
AppOrchid’s unique solution combines a gamified UX approach, powerful predictive analytics and a natural language-based interface that helps the portfolio planner perform what-if analysis of their system.
For example, by simulating variable weather conditions using solar production models and predictive planning, it is possible for managers to estimate the total generation by the electric grid ahead of time. Every 5-minute iteration of solar production data across the grid lead to massive amounts of Big Data, which the system processes in real-time.
When weather conditions cause solar production to drop, AppOrchid apps draw on IoT sensor capability to execute response options to stabilize load, thereby avoiding stress to the grid. After several iterations of this process, machine learning self-calibrates and optimizes the decision response for future actions of the system.
Finally, cognitive reports provide the intelligence to the end user for all potential solar conditions and the associated system’ response across multiple years. This helps the portfolio planner to plan capital projects, pricing programs and customer contracts with a predictability that was previously not possible.
The AppOrchid platform uses Big data and predictive analytics to perform cognitive functions that traditional business intelligence tools did not offer. Specifically, AppOrchid uses data science and analysis to score objects in a knowledge graph, or a system of connected objects that use associative learning to piece together concepts and knowledge. Such a system needs massive parallelization and in-memory computing that only HANA offers. The platform also uses predictive capabilities along with spatio-temporal reasoning to perform brain-like reasoning and inference analysis on real life problems.
There is no competing platform in the marketplace that can perform multi-dimensional queries and graph computations all in-memory with the option to deploy in the cloud.
The AppOrchid solution primarily targets industries that have large amounts of field data and unstructured content. Energy industry with large assets, healthcare organizations with patient records, insurance companies with large amounts of under-writing data, and smart cities are typical candidates for this solution.
The solution works across multiple industries by combining machine learning, natural language processing, and ambient intelligence originating from the IoT landscape. By integrating unstructured data with enterprise applications and IoT devices, AppOrchid sends the resulting Big data through HANA technology to identify patterns, risks and opportunities previously impossible with traditional analytical tools. Root-cause analysis and investigative results are presented back to the user using story-boards or video based narratives, providing insight not possible with conventional visualization tools.
Join the conversation and follow us on Twitter @SAPstartups.