Copyright 2018 - Clinical Commissioning Groups Association

Beyond Limits

Beyond Limits has emerged as the universal leader in "Applied Artificial Intelligence" (AAI) and Cognitive Cloud Computing based on more than 20 years of proven success supporting NASA and the Space program.

 

Launched in 2012, as a Caltech startup, Beyond Limits leverages the expertise, technologies, facilities and ingenuity from the leaders in Artificial Intelligence and Cognitive Cloud Computing from NASA's Jet Propulsion Laboratory (JPL).

 

Historically, nearly every branch of government has come to our team to design and implement solutions for technical challenges that are often deemed unsolvable. To date, our technologies have benefited from over $150 million in investment over many years of Research & Development and practical application to real-world problems as well as problems outside of our solar system.

 

Today, Beyond Limits has continued to enhance and develop these AAI technologies with the foresight to design complex, yet simple and elegant solutions addressing commercial problems in Healthcare, Oil & Gas, Finance, Manufacturing, Telecommunications, Media and others. Our Cognitive Cloud Computing Framework is one step Beyond.

 

Beyond Limits is actively designing and developing products and services for the burgeoning Internet of Things (IOT) market that we call the Universe of Things (UOT). Our claim to this new market naming convention is simply because our solutions can diagnose, predict, act, respond, learn and reason autonomously based on sensor readings and other disparate data streams before the IOT was born. Now, we are applying our capabilities in the UOT across multiple industry segments in which we operate.

 

THE TECHNOLOGY

 

STRATEGIC, OPERATIONAL, AND FULLY-EMBEDDED INTELLIGENCE

 

SHINE enables cost-effective delivery of complex decision support strategies at 500+ million rules/second on a miniature chip, device, code or app.

 

High performance and efficiency is achieved from a unique mathematical transformation based on graph-theoretic data-flow analysis. This enables intelligence to be embedded in light-weight processors with a minimal footprint. This evolutionary technology is now available for mobile applications and 'small things' enabling Universe of Things (UoT).

 

DECISION MAKING NETWORK

 

When every penny, every life, every drop of oil or nanosecond matters, process is key to optimize efficiency, predict and perform to capacity.

 

Sherlock's unique autonomous reasoning solution generates plausible scenarios and automatically changes focus when a dead-end or too expensive scenario appears. Whether you are manufacturing, optimizing upstream operations for oil & gas, servicing the healthcare system or engaging customers... one must minimize variation and risk and maximize efficiency. It is more than just standardization, automation and computation; today is requires real-time interpretation, human behavioral metrics, monitoring, assessing, sensing, diagnosing, prognosing and planning.

 

COGNITIVE COMPUTING FRAMEWORK

 

Sherlock observes, correlates, understands, learns, reasons, generates scenarios and solves simultaneously for deep learning.

 

Our unique, modular and extensible Cognitive Computing Framework based on Sherlock enables our team to create or extend built-for-purpose applications, tools and platforms to support reasoning and distributed intelligence.

 

Solutions based on Sherlock can support Strategic Intelligence to manage, plan, forecast/predict and optimize performance; Operational Intelligence to sense, assess, diagnose, act, respond, learn, teach, enable and Embedded Intelligence to diagnose, control, configure and correct, all requirements addressing highly complex problems.

 

Our combined technologies enable cognitive computing for use in data interpretation and continuous reasoning systems that weave cognitive intelligence into applications and business processes resulting in Cognitive Businesses.

 

Sherlock's reasoning process includes Autonomic Monitoring (AM), which passively monitors and autonomously adapts to unpredictable changes while making decisions on its own. AM continuously adapts to changing conditions and recognizes wrong information. It is light-weight avoiding computational complexities (passively, not actively monitoring) devoting more resources when things become interesting or suspicious.

 

Sherlock's unique Cognitive Engine (Cog-E) is responsible for the higher-order reasoning process. The Inference Engine that reacts to stimulus and the Model-Based Reasoner that provides guidance over processes and goals control the overall synchronization. There is a Short-term Memory and Long-term Memory component encompassing Truth Maintenance guaranteeing logical consistency and coherence of the reasoning process.

 

Sherlock's novel Probe works in cooperation with Cog-E when data sources are insufficient to disambiguate incoherence. Probe dynamically reconfigures to its environment so as to capture more relevant data, ferreting out noise and focusing on the signal(s).

 

BEYOND LIMITS PRODUCTS

 

UNIVERSE OF THINGS™ (UOT)

 

Smart Devices, Smart Cities, Smart Hospitals, Smart Factories, Smart Drilling, Smart Processes

 

Our exclusively-licensed collection of applications and solutions can manage, control, predict, optimize and perform within and across your UOT. Whether in space, underwater or ground, in business; on devices/machines or across devices/machines; our solutions can be fully embedded and scaled to support the UOT. We have the world's fastest and smallest embedded inference engine and knowledge compiler at 500+ million rules per second to monitor any instrumented system and react in real-time.

 

The Universe of Things (UOT) is a collection of applications and capabilities on the Sherlock platform that can be configured to manage, control, predict and optimize performance within and across your connected devices, applications and systems. Whether in space, underwater, underground, on devices/machines or across devices/machines, our capabilities can be embedded and scaled to support the Universe of Things.

 

THE INTELLIGENCE FACTORY

 

Text, Voice or Personae/Avatar-guided Autonomous Intelligence

 

The Factory is a platform for the creation, management and evolution of intelligent agents/assistants/applications to enable customer/patient/consumer/user engagement strategies, training/education and a broad array of customer service processes and functions within knowledge-intensive environments. We have a unique bi-directional language understanding and generation system that uses intent rather than rigid grammar as part of the technology platform. Additionally the Sherlock Cognitive Cloud Computing Framework processes a basic knowledge base to generate an Autonomous Intelligent Knowledge Base (AI-KB).

 

Solutions built on the Intelligence Factory provide multi-modal text, voice, personae/avatar-guided Q & A capabilities that transform a general or specific knowledge repository into an Autonomously Intelligent Knowledge Base (AI-KB) and Intelligent Assistants. The solution continuously learns and reasons, becomes smarter and can simultaneously integrate location, time of day, user habits, semantic intensity, intent, sentiment, social media, contextual awareness and other personal attributes.

 

THE CYBER SECURITY CENTER

 

Identity Theft, Insider Theft, Document, Database, Photo or Network Cyber Security

 

Enhance your security with a unique fine-grain, parsed and distributed information solution that can self-protect, self-destruct, self-discover, self-organize and self-manage. Simplicity and flexibility allows for emergency access but requires 'intelligent' access control through a novel policy-based approach. Data is turned into many self-contained applications placed in enclaves and re-assembled for access through an AI key.

 

A New approach to security is based on opportunistic self-discovery monitoring, bi-modal cognitive based reasoning (what-happens-if and ho-do-you do discovery) and autonomous self-discovery to resolve ambiguity. Its core strength over other approaches lies in its domain-independence with respect to completeness and truth.

 

NetCrypt is responsible for establishing a trusted network communication between pairs of NetCryption nodes. NetCryption is intended to support environments and ad-hoc scenarios in which a trusted third party is not available - for instance, an environment consisting of two large, inflexible organizations. As such, NetCryption enables a simple yet powerful means to establish trust between organizations while sharing as little information as possible. 

 

DocuCrypt will enhance document/data security via “intelligent” encryption and decryption to protect sensitive personal, legal and/or business documents, pictures and databases. This encryption technology is focused on data/documents at rest and enables viewing a DocuCrypt-protected document without fear of hacking.

 

This fine-grain, parsed and distributed information placed in enclaves is self-protected, and requires ‘intelligent’ access through a unique policy-based approach to re-assemble the document/data through an AI-based key. Without the AI key(s), a hacker can only gain access to a very small piece of the entire document rendering the information useless.

 

BUILT FOR DECISION MAKING USING UNPARALLELED REASONING

 

Success is driven by strategic, operational, and fully-embedded cognitive intelligence capabilities

 

Want to make your data sing? Sherlock is the only autonomous cognitive computing framework with bi-modal reasoning – forwards and backwards, just like humans. Sherlock can generate correct hypothetical scenarios even with incorrect, obfuscated or missing data. Our suite of technologies are flexible, scalable, predictable, resilient, secure and simplistic in their individuality along with having a long history of implemented success.

 

Beyond Limits

 

Beyond Limits has emerged as the universal leader in "Artificial Intelligence" (AGI) and Cognitive Cloud Computing based on more than 20 years of proven success supporting NASA and the Space program.

 

Launched in 2012, as a Caltech startup, Beyond Limits leverages the expertise, technologies, facilities and ingenuity from the leaders in Artificial Intelligence and Cognitive Cloud Computing from NASA’s Jet Propulsion Laboratory (JPL).

 

Historically, nearly every branch of government has come to our team to design and implement solutions for technical challenges that are often deemed unsolvable. To date, our technologies have benefited from over $150 million in investment over many years of Research & Development and practical application to real-world problems as well as problems outside of our solar system.

 

Today, Beyond Limits has continued to enhance and develop these AGI technologies with the foresight to design complex, yet simple and elegant solutions addressing commercial problems in Healthcare, Oil & Gas, Finance, Manufacturing, Telecommunications, Media and others. Our Cognitive Cloud Computing Framework is one step Beyond current approaches.

 

Beyond Limits is actively designing and developing products and services for the burgeoning Internet of Things (IOT) market that we call the Universe of Things (UOT). Our claim to this new market naming convention is simply because our solutions can diagnose, predict, act, respond, learn and reason autonomously based on sensor readings and other disparate data streams before the IOT was born. Now, we are applying our capabilities in the UOT across multiple industry segments in which we operate. 

 

The AGI was initially developed to run on small space probe infrastructures while observing, orienting, and solving problems without human interaction. This AGI is superior to other AI offerings in that can generate its own theories and then seek to prove and refine these theories.

 

The AGI is an integrated suite of components:

 

Hunter: A model based system for extracting intent from English using conversational language. Hunter does not require sentences to be grammatically correct or even complete to provide useful and meaningful results. Language recognition and understanding is based on conceptual models of intent rather than rigid and brittle grammar structures. It is capable of understanding conversational, unstructured, and ungrammatical narratives where other approaches fail and recognizes sentence fragments and connects them together. It includes a chaos-folding plausibility reasoner to only generate logically coherent results that contains a“Do What I Mean” capability, automatically correcting for common syntax errors (e. g. spelling, run-ons, missed-used words/phrases, colloquialisms) 

 

SHINE: High Performance inference engine for identification and interpretation of events. It uses a “knowledge base” or collection of rules as its model representation. Extreme high speed with small footprint: 100 to 500 Million rules per second on desktop systems and up to 1.5 Billion rules per second performance on high end processors.

 

Hypothetical Scenario Generator (HSG): this component is commanded by the Inference Engine component to generate logically consistent and plausible scenarios. The Inference Engine makes sure that any scenario generated by the HSG that contains deductions has a transitive closure over a set of its rules to ensure that unrealistic or erroneous scenarios are not generated. The Truth Maintenance (Trump) component is responsible for guaranteeing logical consistency and coherence of the overall reasoning process and any hypothetical scenarios.

 

The HSG provides a general capability to reason in real world situations in the presence of missing or inaccurate data. It provides a core capability for prognostic forecasting and reasoning on developing faults. Ability to predict future outcomes and correlations under real word conditions of obfuscating and miss-information. When applied to future-state prediction problems, it uses the hypothetical scenarios to generate realistic plausible conclusions. When used to provide situational awareness, it will help separate the truth from the lies to generate a usable set of sanitized information.

 

Sherlock: the reasoning engine is a new approach to cognitive computing based in opportunistic self-discovery monitoring, bi-modal cognitive-based reasoning, and autonomous self-discovery to resolve ambiguity. It autonomously shifts through corridors of information to discover plausible facts and scenarios from interpretations of data. It avoids the usual computational complexities of traditional systems by using a technique called autonomic monitoring that enables serendipitous discovery during cognitive forensic analysis. To resolve ambiguity and peruse promising hypotheses it automatically activates user-supplied instruments (tools, probes, etc.) to gather supplementary data to follow a hunch or narrow its scope. It simultaneously and efficiently reasons in both the forward (is this a bomb?) and backward (e.g. Am I being attacked?) directions to tackle a problem from multiple perspectives.

 

Frontier: A general purpose Machine-to-Machine front end for real time standardized interface into arbitrary data sources. It interfaces to real-time data and converts raw data to higher-level data, storing the values in a short-term cache and a long-term relational database. It provides the general information exchange for all the reasoning components and is a proven mechanism for making data available when you need it in the form you need it.

 

 

Content provided by Beyond Limits. For more information please visit www.beyond.ai.  

f t g m