Michigan IT Symposium 2019

When you work with open-source technologies like Drupal, the sky’s the limit. Setting the limit a little lower, so that your team builds solutions that work well within your organization or unit’s digital ecosystem, get done on time, and are things your users actually want, is even better. 

The Web Services team at UM-Dearborn is cross-functional, and our use of Agile development, Lean UX, and open communication have transformed our process, team dynamics, and stakeholder relationships. 

Now, our solutions are effective and efficient from technological, business, and user perspectives. In this session I’ll go over the foundational principles of Agile and Lean UX and bring them together with real-life examples so that we can all be more people-centered in our solutions:

• Make stakeholder kickoff meetings and user research sessions goal-oriented and (possibly) fun.
• Make user stories and requirements more real and generate more empathy within your developers.
• Bring UX and development into the room and make tough choices together, earlier.
• Strengthen your relationships and build trust.
• Make training documentation relevant and clear.
• Reduce rework, leverage tools you already have, and increase the efficiency and happiness of your team.

The University of Michigan is a non-profit institution with a $10B annual budget, 48,000 employees, 46,000 students, and includes a hospital, police force, transportation network, power generation, housing for thousands of students, and more. In 2016, the central Business Intelligence team was tasked by the President of the University to develop University and unit level reports on everything from students, to budget and finance, to HR and research. To accomplish this goal, connections were required across dozens of units and data providers across campus. The ensuing conversations ignited many ideas for improving those datasets and helping units determine what information they both wanted and needed.

This presentation will review how Tableau was used to overcome numerous hurdles, including utilizing dozens of data sets with different naming conventions, different data systems, and other inconsistencies, plus the need for cross-data set calculations. The Business Intelligence team utilized Tableau to create standardize reports accessible via a single and customized platform to deliver a holistic reporting experience for leadership to make better data-driven decisions.

Hear from successful U-M leaders who in their career either have an IT background, are passionate about technology, or hold an IT leadership position. They will share their lessons learned in having successfully navigated the university education landscape while maintaining work-life balance.

Often times when we hear the term mentoring we automatically have an opinion based on our own experience. The presenter will aim to demystify the preconceived notions around mentoring and set participants up for success in their everyday lives on how to be an effective mentor and how to access mentoring. The difference between mentoring and coaching will also be discussed. This session helps participants focus on being intentional about relationship building that can make a difference in career success as you negotiate the workplace.

BIO: Dr. Heidi Sherick offers over twenty-five years of experience working in a variety of administrative roles in higher education including academic affairs, student affairs, and alumni affairs. Currently, Heidi is the Director of Leadership Development in the College of Engineering. Her research involves developmental relationships in higher education and she investigates the processes through which leadership is fostered, including mentoring, coaching, role-modeling, sponsoring, and networking.

The concept of Audiovisual (AV) over internet protocol (IP), or AV over IP, has been around for some time. For the University of Michigan, the idea of using IP in the facilities AV space is new.

AV over IP is the transmission of audio, video and control signals, distributed in real time without latency, over a network cable infrastructure, i.e. WAN, LAN or the internet. In comparison to conventional analogue AV environments, AV over IP refers to the use of standard network equipment to switch and transmit video and audio signals.

In a panel discussion format, AV and IT professionals from COE, ITS, LSA, Michigan Medicine, and Ross will share their experiences of installing AV equipment using AV over IP technology. 

There are many details to be aware of when designing AV over IP-based systems. The panel will share information on how they implemented the technology into their networks, things to look out for, things to consider and references for more information. 

The Duderstadt Center and the Emerging Technologies Group created an innovative learning environment with built in XR technologies. In a Sandbox and Innovation session, we would like to discuss what we have learned during the process of developing the Visualization Studio and the direction we hope to take the space and programs over the next couple of years. 

Health Information and Technology Services (HITS) is a shared services provider of the “core IT” for Michigan Medicine, supporting patient care, research, and education across the U-M Medical School and Health System. On the basis of head count and annual IT hardware/software purchases, HITS represents roughly 2/3 of Michigan Medicine IT.

As a non-revenue generating unit, HITS is charged with stewarding scarce financial resources to optimize for support of daily operations, financial efficiency, life cycle management of IT assets, information assurance (compliance and data security), and flexibility to align with the Michigan Medicine strategic plan.

The composition and financial scope of HITS is dynamic. Drivers such as growth, consolidation of core IT expenses and physical assets, technology change, and new functionality require the Finance and Administration team of HITS to calculate and forecast funding requirements and potential scenarios over a multi-year horizon, as much 10 years into the future. The Strategic Financial Framework (SFF) is a process and a set of templates/tools to “do the math” of creating a common reflection of multiple financial inputs. Further, the SFF exists to spur engagement and participation across HITS leadership. The SFF is a mechanism to “zoom out” and to demonstrate the resource and financial impact of multi-year projects, investment options, or competing priorities. While forecasting is inherently uncertain, HITS seeks to maintain fiscal credibility by carefully weighing and isolating known uncertainty, good faith estimates, and formal multi-year budget commitments to ensure prudent allocation of incremental dollars entrusted to HITS within Michigan Medicine. 

Virtual Care: Successful Telemedicine Programs – Rolling out a successful telemedicine program requires significant collaboration and coordination between people, organizations, and health information technologists.

Connecting MDs to MDs and MDs to patients to optimize patient care, minimize inefficiencies, and maximize access requires orchestration between patients and providers, billing, scheduling, compliance, insurance reimbursement, EHRs/PACs, affiliate/partner organizations, credentialing, and training. Tele-stroke, tele-neurology, pediatric tele-epilepsy, tele-maternal fetal medicine, neurointensive care tele-rounding, Michigan Child Care Collaborative (MC3) pediatric and perinatal tele-psychiatry and child protection team teleconsults are some examples of Michigan Medicine’s telemedicine successes. Also, over 30 different multidisciplinary tumor boards allow radiologists, pathologists, surgeons, oncologists and many other specialists to connect and optimize complex patient diagnosis and treatment. 

This patient care utilizes an infrastructure integrating Polycom, Cisco, Vidyo, Blue Jeans, and peripheral devices to provide connectivity between Michigan Medicine and its affiliates and strategic partners around the world including MidMichigan Health, Metro Health, MD Anderson, Ghana, Ethiopia, India, and many other locations.

To date, the Center for Academic Innovation has developed over a dozen software tools from faculty ideas into full applications, some of which have become commercially available products. With the help of our faculty innovators, we have established an effective process for idea exploration, product planning, software and content development, implementation, and iteration. 

Our presentation will focus on four practices of our team that have contributed to our successful software development: 

• Collaborating closely with a faculty innovator who has deep knowledge of the problem we’re addressing, while our team works to develop a software solution for it
• Focusing on building an MVP (minimum viable product) and then iterating once we learn from the way users engage with the MVP Drawing on the skill sets and perspectives of a multidisciplinary team to ensure that we build well-designed, user-centered software that solves real problems 
• Considering the future commercialization opportunities for a software tool from the beginning of development, so that we build for scale and breadth of applicability
• Through experimentation and iteration, our team continues to hone our approach to evaluating ideas and building software. In this session, we are eager to share what we’ve learned so that others can reflect on our process, pitfalls, and successes in order to inform their own practice.

ArcGIS Online is a simple yet powerful interactive mapping tool to which everyone at the University of Michigan has access. ArcGIS Online can be used to visualize data, analyze spatial patterns, and present materials in a professional-looking app. 

In this hands-on workshop, we will learn how to use ArcGIS Online to easily turn a spreadsheet into a map, discover and add data from authoritative sources to the map, customize the map's appearance, and publish the map for sharing, all on the web. We will also look at some of the options for analyzing and presenting map data, as well as some of the tools and technologies available for collecting geographic datasets.

The My Learning Analytics (MyLA) team will share their experience developing and implementing student-facing visualizations of Canvas course data. With a focus on providing students actionable insights, the team drafted, discarded, and redesigned many prototypes along the way to piloting. This presentation focuses on the impact of collaboration between faculty researchers and ITS staff to design, develop, prototype, and pilot visualizations that support student learning. We hope to give a flavor of the process, the resulting software, and the preliminary research results.   

Partnering with faculty and student researchers from the U-M School of Information and School of Education, a team from ITS Teaching & Learning developed MyLA, a set of student-facing data visualizations about student learning activity. Research on previous student-facing dashboards informed development of the tool. The visualizations provide a transparent view of students’ course standing, reveal behavioral patterns associated with good learning skills, and guide decisions about actions students can take that may improve their academic outcomes. To increase visibility, we embedded the visualizations in the normal learning workflows of the LMS.  

Presenters will share the story of the tool development, as well as lessons learned and feedback from initial student users. Discussion will address how the team leveraged expertise in learning, motivation, and information visualization in an iterative approach to design and development. We will also discuss how this collaboration supports faculty research goals, provides a successful proof of concept as the first application to use the nascent Unizin Data Platform (UDP), and provides a service that supports student success.

The ability to interact with devices through voice and conversational user interfaces is opening many possibilities to enhance and expand existing services. Some talk about voice now similar to the early days of the web, and could be similarly ubiquitous in a few short years.

Staff from ITS Teaching and Learning and Information Quest would like to lead an innovative session discussion around how we've incorporated voice into our existing applications and how they can further enhance teaching, learning and other services.

In order to accommodate a recent growth in enrollment, Engineering has begun utilizing a suite of SAAS applications to: allow for online scheduling (with approval workflow), sync classes from Mpathways into the application, utilize a scheduling algorithm to optimize the class to classroom assignments and generate analytics and reports. These software products allow the College of Engineering to maximize the use of our classrooms while utilizing analytics to make data informed decisions. We have been working with the Provost Office, Registrar’s Office, ITS, and others across campus to implement a solution that could be used by others on campus. 

Participants need to register for Part 1 and Part 2.

Learning how to ask for and offer help has been proven to build positive culture, increase employee engagement, and improve productivity. In this two hour session, attendees can hear from business process owners who will present brief summaries of challenges they face in their departments, including topics such as tracking visiting scholars and managing budget commitments. We will then break out into groups to collaborate with peers on different ways to leverage technology and our experience to solve the issues presented.

Do you ever wonder whether there are others across campus that have similar business use cases where technology is required to address the need? Does it cross your mind that there are probably others who are also defining requirements, working through vendor selection, implementing and then working through how to find a sustainable way to support the very technology solutions you are working on? 

In this session we will discuss best practices and lessons learned from several efforts to build consortiums across campus for applications to meet technology needs. The session will also look at risks and issues to consider, hurdles to overcome, and how to engage others across campus. We will use real life examples of cross campus cooperative projects both completed and in progress, some of which have worked pretty well and others that have encountered issues. Projects examples will include efforts such as Tableau (data visualization tool), Salesforce (CRM), Interfolio (Faculty 360), and Digital Asset Management.  

Come and join the conversation on how together we can leverage our collective resources to address campus wide technology needs, particularly in regard to enterprise level applications.

The University of Michigan generates an incredible amount of data each day, informing ground-breaking research and discovery, teaching and learning, and patient care. With advances in machine learning, advanced sensor technology, robotics, automation, and the Internet of Things, our accumulation of data—as well as the outcomes that can be drawn from it—is expected to grow exponentially.   

Data consumers are often tasked with being experts in databases, data stewardship, and complex analytical tools to access and draw insights from their desired information. In a rapidly changing environment, this limits our ability to make timely and effective decisions. Moreover, our current methods of sharing data often compel campus partners to create copies of institutional data in local databases for individual reporting and decision support needs. This duplication leads to increased IT security, privacy, and compliance risks, poor quality, and different interpretations of the same data. It also creates multiple sources of truth.   

But there is a better way for the future. Join us to hear about some of the innovations that are being pursued to provide easy access to intuitive, consistent and accurate data. We also welcome your ideas on how to advance the use of data at U-M.

The U-M School of Dentistry has employed natural language processing, machine learning, and graph database technology to parse our plain-text patient histories and connect them to the patient's treatment history and a standard tree of health care concepts, the Unified Medical Language System (UMLS). By associating free text with computable knowledge, we have unlocked a trove of previously unusable information about our patient population.

Applications of this tool include identifying communities of patients with similar health and dental care patterns for more targeted precision care, searching for populations of patients for research, identifying gaps in the medical record, improvements to our history-collection processes and forms, and reconciling patient histories at the School of Dentistry with those at Michigan Medicine.

This talk will describe the process and tools used to build the knowledge graph and demonstrate example applications of the tool. We will also discuss our plans for continuing improvement and extension. Technologies used include: Neo4j, MetaMap, MySQL, CLAMP, and Python with scikit-learn.

Are you worried about being sued because your videos don’t have captions? Have you ever wanted a text transcript of a media recording to help with your research or a blog post? Maybe you’re just wondering what all the fuss is about. Don’t know how to get started? If you this sounds like you, we have the answers you are looking for!   

In this hands-on and discussion session we will experiment with free and easy captioning tools, discuss best practices (like what it means to be ADA compliant), and uncover other benefits for captions besides accessibility compliance.

In this interactive session, explore and collaborate with others working to improve business relationships. 

We will cover:

• Good conversations
• Trust
• Meeting people from their perspective
• Good questions for better solutions 

Each section will have an activity and a practice moment in groups of 3 with the roles: observer, relationship manager and client. There will be time in each section for all to collaborate and coordinate best practices.

Participants need to register for Part 1 and Part 2.

Learning how to ask for and offer help has been proven to build positive culture, increase employee engagement, and improve productivity. In this two hour session, attendees can hear from business process owners who will present brief summaries of challenges they face in their departments, including topics such as tracking visiting scholars and managing budget commitments. We will then break out into groups to collaborate with peers on different ways to leverage technology and our experience to solve the issues presented.

Anne and Joe will describe the collaboration efforts on campus for supporting Digital Scholarship and will outline the opportunities for IT partners across campus can get involved. 

Digital Scholarship is often described as being centered in the Humanities, but as being highly interdisciplinary. It commonly involves research that creates or interacts with digital archives. A common challenge with a digital scholarship project is balancing the needs for innovation with the needs for long term preservation.   

At U-M Library, Anne leads the Digital Scholarship team as Digital Scholarship Strategist. At LSA Technology Services, Joe leads a Digital Scholarship Working Group in coordinating, reimagining, and refactoring services to be more supportive of and more compatible with digital scholarship.

With Alteryx, Finance has been able to save hours of time gathering, cleaning and blending together data from various sources and formats. But not only has Alteryx significantly reduced the time-to-deliver data to our customers, Alteryx has enabled new capabilities not previously available. One use case we will present discovered over $200K in duplicate payments to suppliers in the first year of implementation. 

In this session, we will demonstrate the ease of building Alteryx workflows and a couple of the best use cases we have implemented in Finance.

With collaboration from Google representatives, U-M staff from ITS and the Precision Health Initiative created a secure environment modeled after the Google Solution/Architecture: “HIPAA-Aligned Cloud Healthcare” and “Setting up a HIPAA-Aligned Project.” The team provisioned and configured the vast majority of the architecture through Terraform, making sure that the code could be repurposed if any other event or project required a similar setup. The back-end projects are separated by function and IAM permissions are applied accordingly. The infrastructure utilizes Google products such as BigQuery, Google Cloud Storage, Stackdriver Logging, Monitoring, Alerting, and Organization Policies, while the datathon participants primarily utilize Google Datalab and Bigquery to analyze the data uploaded by U-M staff.

The presentation will give an overview of the infrastructure, including design choices and some lessons learned during the process.

Ansible is an open-source automation tool that is being used across the university and worldwide for configuration management and automated application deployment. Automation with Ansible saves time and minimizes errors, allowing our infrastructure to be declarative. We can know the exact state of our infrastructure, how an application was installed, and can rapidly scale resources with minimal direct interaction. 

We have been leveraging Ansible across LSA TS and HITS. In this session, we will share the ways Ansible has helped minimize the time and resources needed for post-build configuration and ongoing maintenance such as software patching. We will provide examples of common administrative tasks that can be simplified, and combined, using Ansible. 

In addition, we will take a deep dive into Molecule, the testing framework for Ansible, and show how it is being used to test configurations, deployments, and software lifecycle paths in local containers and/or virtual machines before impacting real-world systems. We will describe our techniques for debugging and testing using Molecule, while sharing the common challenges we faced along the way.

This session will demonstrate the UDOIT accessibility checker for digital course materials in Canvas, with emphasis on the informational and teaching features of the tool. Possible approaches for introducing faculty to use of the tool will be suggested, and workflows for using the "fix-it" features demonstrated.

This will be a hands-on session using the API directory, including how to create applications, subscribe to an API, and test the applications using UI and Postman.

Nearly eight years ago, a strategic working group was formed to define the Vision for Michigan IT. As the U-M environment changes, the Michigan IT community grows, and we celebrate our sixth annual Michigan IT Symposium, it’s time to evaluate the community’s needs to ensure we are focused on the right goals and objectives. Join your peers for table discussions and exploration of how our community and our programs should evolve to achieve our new U-M technology vision.

Our team undertook the retirement of an 18-year-old application (OSC) written in deprecated code. We would like to present on the uses of Django/Python, why customers should and can use our product, and how customers can join Information Systems and others in ordering and reporting through the SRS.

I never thought of myself as a mentor until I did! When the University of Michigan IT Mentorship program opened up, I applied. This year is my third volunteering as a mentor through the program. Each year is different. I'll describe the learning process, getting the most out of the program while helping your mentees, working on trying to overcome the Impostor Syndrome, and why you should think about becoming a mentor or a mentee.

I encourage an interactive discussion and breakout groups. I'll also ask for suggestions on best mentoring practices, formal, and informal, and being in a dual mentee/mentor role.  

Containerization is a virtualization method offering several benefits over bare metal and virtual machines. As environments are “contained,” each can have specific libraries — or different versions of the same library — that are unique to each application, without conflicting with other apps on the same host. This reduces deployment-stage errors, increases the portability of applications from on-premise to the cloud, between cloud vendors, and more.

Many university groups are using containers for production applications. Who are they and what are some of the use cases they are meeting? This session will be a discussion panel of some of those using containers in production

To ensure that the university has the network it needs to remain a leader in data intensive science, attract the best faculty, students, and staff and enable the next generation of research, ITS Infrastructure Networking group is transforming and revolutionizing the university’s network architecture. The measured annual growth rate in internet traffic of 31% inbound / 22% outbound, means network capacity is currently doubling about every 3 years. Internal flows between data centers for research needs are taxing the limits of the current firewall and deployed IPS solutions.

The networking team is working on deploying network border infrastructure security based on LBL ZEEK; validating different vendor solution for the future core network; reducing manual, repetitive network administration work through automation; deploying network metrics framework and developing and, in collaboration with the academic, research and administrative units, establishing effective network governance processes.

With this presentation, the ITS Infrastructure Networking group will provide an overview of all the in-flight and planned initiatives that will support further advancements of the U-M teaching and research mission.

The Office for Institutional Equity (OIE); the University Library; Information Technology Services (ITS); the College of Literature, Sciences, and the Arts (LSA); and other campus partners are piloting the Universal Design Online Content Inspection Tool (UDOIT) in Canvas, in order to address more content accessibility issues earlier in the course creation process. UDOIT, in conjunction with the Canvas learning management system, can help to inform institutional strategy and best practices related to making learning materials more accessible for students with disabilities.   

Every term thousands of documents are uploaded to the university’s learning management system by faculty for students to read as part of their coursework. Many of the learning materials go unexamined for accessibility barriers. UDOIT provides an automated solution for ensuring that common pitfalls are caught and corrected, or simply avoided (e.g., missing alternative text for images or well-formed HTML pages for screen reader support). While faculty generally seem interested in ensuring their course materials are accessible, we face the challenge of scaling meaningful support for faculty when proactively doing accessibility work.   

Our goal is to establish a higher standard of accessibility in learning materials at U-M so that students with disabilities face less barriers to learning. Localized training materials will be created that will support faculty and staff self-service use of the tool. A pre and post scan of a representative sample of select courses will be conducted. Data from the scans will help inform our strategy for addressing accessibility in courses across the institution.

Research takes place all over campus. For the ISR, it’s our primary mission. As a result, we are responsible to both our funders and faculty with what happens because of that research. Many of our grants require us to track publications that their money/data produced, both inside and outside the Institute. Currently dozens of departments individually track those publications, and there is plenty of inefficiency due to the overlap of those departments’ faculty. 

 This project has three objectives:

1. Allow for an easy ingest, via manual and/or automatic means. 
2. Create a review process for duplicates and annual review for faculty
3. Generate useful output for reporting (faculty(e.g., CVs) and funders(e.g., data publication lists)). 

Tracking does not stop at the walls of the ISR though, but goes with access to our data into the world. Not only can this project benefit the ISR, but the U-85M academic community as a whole by creating a cohesive, accessible environment for the broader use of everyone. Building this as a tool that benefits everyone requires input from multiple areas, and I believe the symposium’s Sandbox & Innovation Session provides the perfect environment to make this discussion effective and relevant.

Have you ever wondered what the Michigan IT Hacks with Friends event is all about? It is a special event each year where designers, developers, documenters, and supporters get together and build prototypes of applications that solve problems. Do you have any ideas for applications that would improve your work life, helping with the mission of supporting research, teaching, or administration? Please join us for this break-out session, where we will remagine and reinvigorate the purpose and structure of the Hacks with Friends event.  We will discuss how we can be more inclusive of participants from all members of the U-M community, as well as potential changes to the name of the event, categories of prizes awarded, and how pitches are conducted prior to the event.