This thesis is about the collaboration and interaction between a robot and a human agent. The goal is to use the robot as a coworker, by implementing the premises of an interaction system that would make the interaction as natural as possible. This involves that the robot has a vision system that allows understanding of the intentions of the human. This thesis work is intended to be part of a larger project aimed at extending the competences of the programmable industrial robot, Baxter, made by Rethink Robotics. Due to the limited vision abilities of this robot, a Kinect camera is added on the top of its head. This thesis covers human gestures recognition through the Kinect data and robot reactions to these gestures through visual feedback and actions.

Varför behöver tillverkningsprocesser automatiseras? Anledningarna är flera, delvis så är det ettsätt för högavlönade länder att konkurera med lågavlönade då det kostar lika mycket att investerai tekniken i båda länderna. Det kan också underlätta tillverkning av komplexare produkter medmindre felmarginal.Safetrack som är ett företag som specialiserar sig inom kabelanslutningsmetoder och järnvägsmaterial. En av deras tillverkningsprocesser är i behov av automatisering. Processen sombehöver automatiseras är tillverkningen av lodpinnar med tråd. Lodpinnar med tråd befinner sigpå en nedåtgående marknad, detta betyder att Safetrack inte vill göra någon större investeringför dessa produkter. Idag så görs allt mer eller mindre manuellt med hjälp av semiautomatiskamaskiner som pelarborr och pneumatiskt styrda skruvstäd.Syftet är att undersöka hur denna tillverkningsprocess kan underlättas genom automatiseringav vissa moment i denna process. Målet är att ta fram ett koncept till Safetrack på en automatiserad tillverkningsprocess som är effektivare och tidbesparande jämfört med den nuvarandeprocesssen. Metoder inom teknisk problemlösning enligt Ulrich och Eppingers bok “Produktutveckling: Konstruktion och Design” används för att välja vilket koncept som ska utvecklas. För att ett koncept ska kunna utvecklas behöver först och främst automationsnivån bestämmas. Detta görs genom att identifiera vilka av de nuvarande stationerna som är i störst behov av automatisering. Automatiseringgraden som beskrivs av enligt metoderna i “Smart Automation” och kallas LoA(Level of Automation).Koncept för två maskiner tas fram. LoA höjs till en lämplig nivå på alla stationer som utvecklas.Den första är en helautomatisk maskin som klipper och knorrar koppartråden. Den andra är entrådfästningsmaskin som med input från operatören fäster tråden i lodpinnen.För att konceptet ska kunna förverkligas så måste CAD-modellerna till båda maskinerna till-verkningsanpassas till lämpliga tillverkningsmetoder. Innan produktionen påbörjas måste det säkerställas att CE-märkning uppnås för båda maskinerna. / Why are manufacturing processes in need of automation? One reason is that it lets countries where employees have higher salaries compete with countries where employees have lower salariessince the cost for investing in technology is the same in both countries. Another reason is thatautomation allows for manufacturing of more complex parts with a smaller margin of error.Safetrack is a company that specializes in modern cable application methods and equipmentfor the railway industry. The manufaturing process for one of their products is in need of automation. The product is called brazing pins with fuse wire. The product is part of a declining market,because of this Safetrack doesn’t want to make too large of an investment. The manufacturingprocess is more or less done manually with the occasional help of semiautomatic machines suchas a pillar drilling machine and pneumatically driven vises.The purpose of this study is to examine how certain steps in the manufacturing process can bemade easier through automation. The goal is to develop a concept for an automated manufacturing process that is more time-efficient than the current process. Methods on technincal problemsolving will be used according to Ulrich and Eppingers book “Product design and development” to choose which concept to develop further. For a concept to be developed the level of automation needs to be determined. This is done by identifying which parts of the current method are in the most need of automation. The level of automation is described in“Smart Automation” and is abbreviated to LoA.The concepts for two machines are developed. The LoA has increased for both of the machinesthat are developed. The first machine cuts and shapes the wire entirely automatically. The secondmachine joins the wire and the brazing pin with the input from a worker.For the concept to be made a reality all components must be adapted to fit a chosen manufacturing process. For the machines to enter production they must also fulfill the requirements forCE marking.

Modern synthetic gene regulatory networks emerge from iterative design-build-test cycles that encompass the decisions and actions necessary to design, build, and test target genetic systems. Historically, such cycles have been performed manually, with limited formal problem-definition and progress-tracking. In recent years, researchers have devoted substantial effort to define and automate many sub-problems of these cycles and create systems for data management and documentation that result in useful tools for solving portions of certain workflows. However, biologists generally must still manually transfer information between tools, a process that frequently results in information loss. Furthermore, since each tool applies to a different workflow, tools often will not fit together in a closed-loop and, typically, additional outstanding sub-problems still require manual solutions. This thesis describes an attempt to create a tool that harnesses many smaller tools to automate a fully closed-loop decision-making process to design, build, and test synthetic biology networks and use the outcomes to inform redesigns. This tool, called Phoenix, inputs a performance-constrained signal-temporal-logic (STL) equation and an abstract genetic-element structural description to specify a design and then returns iterative sets of building and testing instructions. The user executes the instructions and returns the data to Phoenix, which then processes it and uses it to parameterize models for simulation of the behavior of compositional designs. A model-checking algorithm then evaluates these simulations, and returns to the user a new set of instructions for building and testing the next set of constructs. In cases where experimental results disagree with simulations, Phoenix uses grammars to determine where likely points of design failure might have occurred and instructs the building and testing of an intermediate composition to test where failures occurred. A design tree represents the design hierarchy displayed in the user interface where progress can be tracked and electronic datasheets generated to review results. Users can validate the computations performed by Phoenix by using them to create sets of classic and novel temporal synthetic genetic regulatory functions in E. coli. / 2016-12-31T00:00:00Z

The purpose of this thesis is to discuss the design and development of a platform used to automate a stock 2013 Ford Focus EV. The platform is low-cost and open-source to encourage collaboration and provide a starting point for fellow researchers to advance the work in the field of automated vehicle control. This thesis starts by discussing the process of obtaining control of the vehicle by taking advantage of internal communication protocols. The controller design process is detailed and a description of the components and software used to control the vehicle is provided. The automated system is tested and the results of fully autonomous driving are discussed.

Colony Collapse Disorder (CCD) has been a major threat to bee colonies around the world which affects vital human food crop pollination. The decline in bee population can have tragic consequences, for humans as well as the bees and the ecosystem. Bee health has been a cause of urgent concern for farmers and scientists around the world for at least a decade but a specific cause for the phenomenon has yet to be conclusively identified. A normal hive inspection can be very disruptive for the bee colony, as the hive needs to be disassembled to visually assess hive health from the inside by collecting larvae and egg data. This work uses Machine Learning and Computer Vision methodologies to develop techniques to monitor hive health without disrupting the bee colony residing in the hive. Bee traffic refers to the number of bees moving in a given area in front of the hive over a given period of time. Bee traffic is related to forager traffic. Forager traffic is the number of bees moving out of the beehive. Forager traffic is a crucial factor in determining and monitoring food availability, food demand, colony age structure, the impact of pesticides, etc. on beehives. This work focuses on estimating bee traffic levels in a given hive and associate this information with data collected through manual beehive inspections.

No description available.

With autonomous vehicles becoming more common and established, there are some problems to overcome before their full potential can be reached. One of these problems is the lack of information about the condition of the road, which traditionally would be acquired from the driver operating the vehicle. Volvo Autonomous Solutions are developing an autonomous hauler, made for operating in off-road workplaces, such as quarries and mines. In these off-road workplaces, road maintenance is limited and often performed only when deemed necessary by a driver. This thesis investigates the issue of detecting irregularities in the road on an autonomous vehicle. To achieve this data from an Inertial Measurement Unit (IMU) and a Global Navigation Satellite System (GNSS) unit mounted on the vehicle is collected, analysed, and classified to find any irregularities in the road. In order to improve confidence in the classification of the irregularities, false positives are reduced by using an occupancy grid solution. The results show that the use of IMU data can be used to detect irregularities and that the use of an occupancy grid increases the confidence of detected irregularities.

A lot of features that need to be scanned underground should not be disturbed, from waterlines to unmarked graves. A non-invasive way of probing underground is Ground-Penetrating Radar (GPR). GPR finds differences in materials with radar waves. However, GPR is human-operated and its position is generally determined with a GPS. In some cases, the presence of a human operator can be dangerous, and in other cases, the GPS is not reliable (i.e. mines, glaciers). Therefore there are situations where an autonomous and non-GPS-reliant solution is preferable. The current state of the autonomous GPR system targeted in this work has a non-intuitive GUI that requires an experienced hand to operate. I present an updated hardware and software platform with an intuitive GUI. This updated autonomous system continuously builds a map of its surroundings with Simultaneous Mapping And Localization (SLAM). SLAM localizes itself within the map through sensor-fused position estimates. After the survey is completed the positions are saved and integrated with radar data to be visualized. Robot Operating System 2 (ROS2) is the software I used that allows communications between hardware components, software systems, and the GUI. The new hardware package uses only one source of power and is built using quick connectors that allow for quick removal from the GPR platform. This system allows for intuitive autonomous survey planning and execution in any field paired with a simple way of visualizing data.

This exploratory case study examined how and if career counselors learned about the impact of automation and artificial intelligence (AI) on occupations, including which actions, activities, and conditions helped or hindered their learning, and what impact this learning had on their professional practice. The data included questionnaires and interviews with 15 participants who worked as career counselors and expressed that learning about the impact of automation and AI on occupations was important in their work. Three primary findings emerged from the study. The first finding illustrated that participants primarily learned through informal and incidental learning and all participants described the lack of a clear path to learning. This was framed through four subthemes, including learning that emanated from reflection on “Aha!” Moments, learning that involved a process of continuous learning, formal learning that was lacking but desired, and participants stating that they did not seek out the learning. The second finding revealed that agency, openness, collaboration, and reflective practice helped learning, while limits to time, resources, and opportunities for learning hindered learning. The third finding showed that such learning impacted participants’ professional practice by affecting their interactions with clients and their professional development as career counselors. This demonstrated the application of the learning and connected it directly to the work of career counselors. This study contributes to the growing body of literature examining informal and incidental learning in complex and chaotic domains and provides recommendations for research and practice, specifically addressing how the results of this study can be used to support the professional development of career counselors in the Fourth Industrial Revolution.

No description available.