The AI World Cup competition is a once a year event to develop competitive, adversarial AI technology that can directly compete against other teams. The challenge is to use various AI techniques to create a competitive soccer team to compete against others. This is a very good learning competition to learn various types of situational AI which can be applied to various real world, adversarial problems. The website is http://aiworldcup.org/

The second AI World Cup 2019 is scheduled to take place on October 31th – November 2nd, 2019, on KAIST Campus, Daejeon, Korea. This is the first world‐wide competition of on‐line soccer teams simulated by AI. The aim of this competition is to bring together researchers and developers of AI technology, both from academia and industry, and to prepare solutions involving AI technology for a rapidly evolving world. Participants will be provided with a package of instructions on how to use (on‐line soccer players) and API for coding their game plans. Tutorial lectures and manuals with detailed instructions are available on the official website. All participants are requested to enter qualifying round(s). Only qualified participants can advance to the main round, hosted on KAIST Campus on October 31th – November 2nd.

• AI World Cup Competition ( 2019 )

The Agricultural market sector is one of the most promising areas to reap the benefits of IoT technologies. This project will examine the IoT ecosystem and apply it in an agricultural environment. From the various sensors used to monitor environmental conditions, to the complex networks used for data transport, to the systems used to store and analyze the large information collected, we will develop a prototype system for crop monitoring and feeding. Particular emphasis will be placed on analyzing, implementing, and comparing the various wireless networks that are emerging in this area for sensor connectivity and aggregate transport of data to offsite locations. After building a proof of concept system, time permitting, we will implement it on an actual working farm. Alternative rural energy systems may also be explored and implemented to power the system at remote locations.

• Construction of Farm-related Networks and Data Management Using LoRa Protocol ( 2019 )
• Development of Eco-friendly Irrigation System Using LoRa ( 2019 )

Build a completely autonomous drone requiring NO RF transmission or receipt to be able to fly autonomously on a pre‐programmed course. Be invisible and invulnerable to RF based CUAS systems. This is a challenge project for development of red team/blue team counter UAV development.

• Augmenting Structure for Noise Reduction in UAV ( 2019 )

Hi Tech Scarecrow is a startup company with the intent to serve the agricultural sector. A main issue of crops, especially high value crops such as tree fruits, is the damage done by birds to the fruit while still on the tree. A scarecrow is an old notion of a human like figure that will scare away birds from the trees. Hi Tech Scare Crow is a modernization of this concept using equipment such as robots, UAV’s, sensors and different machines to drive birds away from a fruit orchard.

Wea Creek Orchard is a 6‐acre U‐pick apple and peach operation on the south side of Lafayette. During ripening periods for our apples (and to a certain extent for our peaches also), we have problems with birds perching on the fruit and pecking away at them. Obviously, these fruits are then unsellable. Last year we lost about half of our crop of the largest red apple which is quite popular among u‐pick customers. Each year’s loss is not that great in the scheme of things, but I would guess tangible losses last year was in the $100’s, and this added up over time show that something needs to be done. A corollary problem is deer foraging on the tree bark, leaves, and fruit throughout the year. If these problems could be solved simultaneously we could save profits approaching $1000’s. Having a dynamic system that pests would be less likely to be habituated to would be ideal.

• Deep Learning-based Bird Sound Recognition System With Data Pre-Processing ( 2019 )
• LoRa based Bird and Deer Repulsion System in Farm ( 2019 )
• Dynamic Bird Detection Using Image Processing and Neural Network ( 2019 )

The Internet-of-Things (IOT) is one of the fastest growing technologies evolving to enable new human technologies today. In this project, the team will develop applications for IOT and the infrastructure for IOT for a specific area. In this case we will design an infrastructure to perform management tasks on the Korean Software Square (KSQ) so it is more “self managing”. This will include cleaning robots, temperature and humidity control, human-building, human-robot and human-sensor interfaces, security and any other ideas for making the KSQ an intelligent atmosphere.

• Internet Of Things Infrastructure for a SMART Building ( 2017 )
• A System for Determining Braking Distance by Predicting Weight Using Image Processing to Predict Illegal Red Light Running Events ( 2019 )
• Red Light Running Detect System ( 2019 )
• IoT Sensor Based Emergency Detection App ( 2019 )

Despite what appears (and is marketed) to be ubiquitous wireless connectivity across the United States and most parts of the world, there are still large swaths of areas that do not have wireless connectivity and services. In particular, many remote rural areas and areas with challenging terrain are often where communications and location services are perhaps needed the most, but have the weakest coverage or no wireless systems at all. We will analyze the various requirements needed to implement a rapidly deployable system that can provide basic communications and data network needs in an environment typically used in an emergency situation (such as a wildfire) or an adventure racing environment where competitors may be heading into remote, mountainous or desolate area where they need to be tracked and monitored. We will attempt to build and test a rapidly deployable system to address these needs, based on various wireless networking technologies.

• Measurement of indoor and outdoor performance of wireless communication using LoRa ( 2018 )

This project is a part of a development of a small, inexpensive, and portable radar system to detect unmanned aerial vehicles and track trajectories of them for surmising their purpose of flight. One of the most important part to detect and track those UAVs are a monitoring system which displays data transmitted from radar in real time to assist users/surveillance to collect information of moving objects in the surveillance area. The monitoring system that described ahead will be composed with an android device and an application which will be loaded on the android device. By designing the monitoring system using android based systems, the small, inexpensive, and portable radar system provides accessibility to people. The whole surveillance system including radar and monitoring device will allow various advanced researches because of the functionality and accessibility. Key functionality that the application has to have is to capture data through rabbitmq and display the data as a real time graph on the device. Project participants will mainly work with android application development. Given data for this project will be signal-processed data, so that the data domain will be time-frequency or more refined form. The data need to be displayed in timerange form in graph (this can be changed in future), range and angle need to be presented on the screen, and the data need to be stored in a table or equivalent data form. Any idea of the display and data storage are welcome and encouraged.

• Detection and tracking of drones using radar ( 2018 )

The single biggest constraint to UAV usage is endurance. This effort will tradeoff free flight maneuverability for endurance by tethering a powered UAV to the ground to provide power, and allow it to hover, in low to moderately windy conditions, at a fixed point and relay video data of the surrounding ground. The purpose would be to provide a “high view” of an area by placing a camera at an appropriate height (100 ft) to show the surrounding area. The UAV, by being tethered, can hover indefinitely but would have no ability to maneuver; it would be designed to do station keeping at altitude only, but for an unlimited period of time.

• Low Cost Tethered Unmanned Aerial Vehicle System ( 2017 )
• Detection and tracking of drones using radar ( 2018 )

The world is becoming a linked entity, not just for people. HARMS stands for Human, Agents, Robots, Machines and Sensors. This is a large on-going project at Purdue to develop a 5 layer model which can indistinguishably link humans and non-humans together in new, flexible ways. The model has 5 layers; networking, communication, interaction, organization and collective behavior. Participants will work on specific projects linking these actors together in applications for use in real situations.

• Design and Implementation of Unmanned System Restaurant using HARMS ( 2018 )

The growth of UAV problems has opened a new field of Counter UAV Defense. One of the main tasks in defense is the detection and categorization of UAV’s which may be threats. In this project, the team will take a current acoustic node and make it mobile. This will be creating a prototype of a new acoustic node and then testing it in the field with UAV’s

• Convolutional Neural Networks for Analyzing Unmanned Aerial Vehicles Sound ( 2018 )

One of the computational areas currently receiving significant attention from researchers and practitioners is the field of data integration and analytics known colloquially as Big Data, and of particular importance to Big Data projects is information visualization. On said projects, information visualization solutions provide the main mechanism for users to consume, interpret, and communicate the findings from the analyses enabled by Big Data, and hence they are critical to the overall success of the projects. Moreover, visualization can serve as a powerful resource to users in that the opportunity to “see” the data may present dimensions and lead to inferences that otherwise would not be accessible to the users. These users come from a variety of domains and disciplines, and among these is the area of cybersecurity. From a cybersecurity standpoint, one of the most vibrant and compelling Big Data applications is the analysis of network traffic data that serves as a very common and more significantly critical activity in postmortems of cyber attacks as well as efforts to monitor, detect, and even predict malicious activity. Only does network traffic data provide dimensions of volume and velocity but also variety in that complementing network data with additional information such as geospatial and temporal characteristics often render findings otherwise not salient. One promising area related to the analysis of network traffic is community detection. As a well known research area in graphs and networks, the notion of communities has been leveraged in several contexts including cybersecurity (Hogan et al., 2013; Joslyn et al., 2013). However, with newer and faster approaches to detecting communities in networks (Lu, Halappanavar, Chavarria-Miranda, Gebremedhin, & Kalyanaraman, 2015; Lu, Halappanavar, & Kalyanaraman, 2015; Rahimian, Payberah, Girdzijauskas, Jelasity, & Haridi, 2014), it is an open issue whether such approaches qualitatively improve the detection of relevant communities for cybersecurity purposes, and a fundamental limitation to addressing this particular issue is the lack of tools and resources available to researchers that provide agile capabilities for analysis.

• Packet Visualization Tool For Individuals ( 2018 )
• Analysis of the Relationship Between Server Location and RTT at Client Level ( 2018 )

Perceive is helping retailers measure customer influences in brick and mortar stores using video systems. Retailers and marketers today have a lot of information about what their customers do in the online realm before a purchase but they have very little of this information in the digital realm. To change this we are building video devices for retailers to put in stores, streaming the data back to our servers, and analyzing it to understand what customers are doing in the store.

• Video streaming with encryption on mesh network that using batman ( 2017 )

AccuPS is a technology startup developing natural and immersive 3D man-machine interaction solutions, particularly for virtual reality (VR) applications. We are looking for talented mobile app programmers who can develop 3D virtual reality apps running on mobile platforms. The programmers will be utilizing a mobile VR headset and AccuPS’s high precision head/hand-tracking device.

• Screen Sharing With Networking in 3D Virtual Reality Games ( 2017 )

The paperclip fitness platform is a software system to help recreational facility staff with customer outreach and communication, and to create a safe and welcoming environment for their patrons. It includes a mobile app for patrons, a tablet app for staff and a web application for managers and decision makers.

• Fitness Management Platform Using Real Time Database ( 2017 )