Local Theory, Local Practice 

Tri-Citizens sometimes can't tell that they live in an innovation hub. The growing world of technology rarely touches day-to-day life, but we're here to change all that. While cutting-edge technology might be just down the street, often there aren't enough street smarts to truly sell the technology. That changes now.

This year Launch Weekend will include technologies from Pacific Northwest National Laboratory (PNNL) and Washington State University (WSU) to offer a unique private-public partnership to make new ideas. Combining millions of dollars in R&D with local entrepreneurial spirit, the founders of Launch Weekend are hoping the combination will lead to greater innovation. 


SilentAlarm: Malware Beware (PI: Joel Doehle, PNNL)

Adversaries are constantly developing new malware to attack computer systems. Many of today’s cyber security solutions will not detect these novel attacks because these defenses are designed to look for signatures of known malware. SilentAlarm goes beyond signature-based detection to identify adversarial behavior in a network based on abnormal traffic generated by malicious software. This patented technology uses inference-based reasoning to correlate distinct events across a network to pinpoint and subdue potential malware. At Launch, you’ll see how SilentAlarm reasons about events that it is observing and takes action to cut off an offending machine when it concludes that it has been infected. This software is continuing to be enhanced, and we are seeking partners to transition it to commercialization.

StreamWorks: Finding Patterns in Streaming Data (PI: Sutanay Choudhury, PNNL)

Organizations are increasingly looking for ways to detect precursor events and patterns as they emerge. StreamWorks looks for patterns in near-real-time streaming data. This patent-pending technology enables analysts to visually specify a pattern they want to detect in a data stream, then examine the resulting patterns as soon as they happen. When compared with competing products, test results with a particular Internet traffic flow dataset showed up to 100 times improvement in runtime when StreamWorks was used. Launch participants will see how patterns are queried and generated. We are seeking collaborators to pilot this technology for real-time monitoring in cyber security, finance, and Internet of Things applications.

Clique: Online Threat Detection; Detects and Analyzes Threats in Real-Time (PI: Daniel Best, PNNL)

Clique is an advanced data-intensive visual analytic software package that combines visual identification and investigative discovery—enabling detection and analysis of cyber threats in near-real-time. Network defenders now have a mechanism to move seamlessly from high-level views of behaviors down to detailed representations. You will see two views of Clique: 1) Cadence, with a graphical user interface that gives users the ability to see deviations from expected activity, and 2) Trace, which gives analysts a flexible and scalable two-dimensional scatter plot, revealing patterns in large volumes of network data.

Scalable Reasoning System: Analytic Framework for Web-Based Visualization (PI: Scott Dowson, PNNL)

Scalable Reasoning System is an analytic framework for developing web-based visualization applications. Using a growing library of both visual and analytic components, custom applications can be created for any domain, from any data source. Its modular architecture helps connect data to analytics and visualizations—helping users make sense of data. This technology has been used to create solutions for a wide range of domains including health care and cyber security, incorporating either large or streaming data sets.

Creation of Smart Home Technologies to Support Adults with Cognitive and Physical Limitations (WSU Pullman)

In today’s connected world, businesses and individuals are more interested in capturing and understanding the data our smart phones, smart speakers, even smart refrigerators transmit. This data‐mining craze has created a proliferation of sensor data and we are finding it difficult to track all of these smart devices in a given environment.

This WSU invention is a smart home, named CASAS that uses artificial intelligence technologies to learn a profile of resident activities in a smart home. CASAS uses the learned information to analyze resident behavior, predict and identify activities, perform automated assessment of resident functional

independence, and assist by automation activities in smart homes. Recent advancements in supporting fields have increased the likelihood that smart home technologies will become part of our everyday environments. However, many of these technologies are brittle and do not adapt to the user’s explicit or implicit wishes. Here we introduce CASAS, an adaptive smart home system that utilizes machine learning techniques to discover patterns in user behavior and to generate automation polices that mimic these patterns. The unique contribution of this work is the ability of CASAS to keep the resident in control of the automation. Users can provide feedback on proposed automation activities, can modify the automation policies, and can introduce new requests. In addition, CASAS continuously updates its models to reflect changes in resident behavior.

Apple Harvesting Machine for Formally Trained Orchards (WSU Pullman)

Improved mechanical harvesting techniques for apples and other specialty crop orchards Harvesting labor accounts for 50% of the total tree fruit production cost and this is expected to continue to rise as labor shortages increase. This presents an urgent need for improved mechanical harvesting techniques in apple and other specialty crop orchards. In addition, growers are adopting new horticulture practices that allow for tree limbs to be trellised to a post or wire because, among other benefits, these trellised systems will be ideal for automated and mechanized harvesting techniques.

Researchers at Washington State University’s Center for Precision & Automated Agricultural Systems have invented an apparatus for bulk harvesting of apples and other fruit on trellised trees such as fruiting walls. The apparatus has a lifting arm which manipulates a catch frame as well as an apple harvesting attachment. The catch frame consists of multi‐layer catch pans, which are inserted into a tree canopy wall and an implement arm which supports the harvesting attachment that engages the apples. A Dual Motor Actuator (DMA) coupled to support shafts is utilized to provide a shaker harmonic pattern and energy to the trellis wires or branches of the trained trees in the orchard. Finally, the apples are removed and caught in the catch pan(s).

Protein-based Superabsorbent Hydrogels (WSU Pullman)

Researchers at WSU have produce a novel superabsorbent material prepared by a macromonomer approach which provides for a superabsorbent hydrogel. Additionally, due to the method of manufacturing the hydrogel is biodegradable.

Superabsorbent hydrogels are water‐absorbing polymers that can absorb 10‐1000 times their weight in water. Currently, superabsorbent materials are mostly made of petroleum‐based polymers, mainly copolymers of acrylic acid (or its sodium salt) and acrylamide. Which can not only be toxic to the environment, but also require specialized manufacturing methods to produce. This technology by comparison, is a protein based product that will reduce environmental impact, as well as fulfill society’s sustainability goals of producing more products from renewable resources. Additionally, the degradation of the gel as well as the mechanical properties can be controlled by varying the composition of the components. Making this technology widely applicable to a number of industries

System and Methods for Nutrition Monitoring (WSU Pullman) 

Diet is an important lifestyle and behavioral factor in self-management and prevention of many chronic diseases. Currently, technological approaches for nutrition monitoring either use computer vision techniques for image-based diet assessment or mobile apps for maintaining electronic food diaries. These approaches are lacking in accuracy, ease-of-use, and privacy and thus adherence to utilizing these technologies in clinical and real-world settings is quite limited.

Researchers at Washington State University’s School of Electrical Engineering and Computer Science have developed a system and associated algorithms and techniques for nutrition monitoring which employs speech recognition and natural language processing for measuring the nutrient information of food from spoken data. A user can essentially speak the name and portion size of the food they ate, and the speech is processed in real-time to compute the calorie intake. As we move into more speech-based user interfaces in mobile devices, this method allows user input in a more pervasive way while increasing the accuracy of calorie calculation.

This system also takes advantage of the United States Department of Agriculture’s (USDA) national database to compute the nutrient per serving calculations. This is an advantage over current calorie monitoring apps which utilize crowdsourcing mechanisms to calculate caloric intake which can ultimately lead to many inconsistencies.