Dates:         April 24, 2012 - 9 am to 4 pm

                    April 25, 2012 - 9 am to noon

Purpose

1.  
2.  
3.  
4. Apply the model to current expansion scenarios including; Montana/Wyoming, TAACCCT grant proposal, HP Catalyst iLab Central project, Alberta Grand Prairie College, BC Shared Service, … 
5. In light of all of the above review and revise NANSLO Scale Network Template and How-to Adoption Manual documents

Boulder Meeting Notes

Expansion opportunities outside of North America:

-          China

-          India

-          Europe

-          Australia

Create a non-profit that represents NANSLO:

-          funnel funding

-          centralized scheduling

-          configuration control (manage different interfaces at different institutions)

-          core set of experiments

Location:    WICHE Learning Centre

                    3035 Center Green Drive, Suite 200

                    Boulder, CO 80301-2204

                    Directions from Denver International Airport

                    http://www.wiche.edu/directory/location.

Attendees:      Rhonda Epper, Pat Shea, Catherine Weldon, Daniel           Branan, Ron Evans, Albert Balbon, Paul Stacey,
                    Scott Seville

Key Questions

Pre-Boulder Meeting Answers So Far

Boulder Meeting Notes

Configuration

What is the minimum configuration of an RWSL node?

-          Physical space

-          Hardware

-          Software

-          Network

What is the recommended configuration of an RWSL node?

What are the configuration options associated with an RWSL node?

-          academic domain specialization

-          all souped up version

Definitions: https://nanslo.pbworks.com/w/page/43653937/RWSL%20and%20RDN%20Definitions

RWSL Node Technology: https://nanslo.pbworks.com/w/page/48382359/RWSL%20Sub-modules

Infrastructure Required: https://nanslo.pbworks.com/w/page/47284100/RWSL%20Node%20Requirements

How-to Adoption Manual

https://nanslo.pbworks.com/w/page/42587037/How-to%20Adoption%20Manual%20-%20Home

Definitions need work. Node = host organization. Lab = equipment for a particular lab. Lab exercise = what you do with the equipment ie. the actual experiment. Need to define set of core experiments.

To get started need:

1. network switch (8 ports/reuired lab module base) with 1 Gigabit network and IT cabling

2. required lab module base equipment

3. lab sub-module related to what lab you want to do

-- biology - microscope

-- chemistry & physics - spectrometer

reccommend two of everything if going with operational use with students. One is ok if just exploring

microscope recommendation would be different if had druthers - current microscope is not reliable (China) and no local support. Would prefer Nikon (maybe corporate sponsor?)

Need to develop standardized configurations for each lab so that documentation and support is congruent. Need to establish/enforce standards to keep installations aligned.

Need a physical space ideally already equipped as a lab (fume hoods, water, etc.) Servers and lab equipment need to be virtually separated from the rest of the network and outside the firewall.

Key Questions

Pre-Boulder Meeting Answers So Far

Boulder Meeting Notes

Staffing

What are the staffing resources required for setup and operation of an RWSL node?

-     experiment design/developer

-     RWSL lab manager

-     RWSL lab technicians

Staffing:
https://nanslo.pbworks.com/w/page/50967107/Tech%20Staff-Support

Skill sets (need to create list of tasks without necessarily specifying roles)

- master level establish set of configuration standards, design labs & expts

- one person (or faculty members) who understands the science of the expt

- lab director setup of node

- need lab techs if using with students

To get started need one person half time.

Capacity

What is the capacity of an RWSL node?

-          # of students

-          solo or groups of students

-          time per experiment

operating hours

Capacity Estimator Spreadsheets: https://nanslo.pbworks.com/w/page/43653060/Scheduling%20-%20Home

Ron: We've already done estimates of how many students a node can handle when we were writing the original proposal.  We know how many hours there are in a week and can estimate how many hours per week a node is likely to be operating, we can estimate how many 'down' hours are expected in a week, and we can estimate how many students/hour (~1 student/hour) a node can handle.  From this information we can estimate how many student-hours per semester we are likely working with.  Each node that comes on-line will come with a certain amount of 'home institution' student hours that will use it before it can serve remote lab services (RLS) to RDN.  Subtracting the number of home institution student hours required from the total number of student hours a node has available will tell us how many student hours that node is likely to be able to serve RLS to RDN if it is required.  I expect this will vary widely from institution to institution.

- 1 standard module

- 2 hours sufficient as a block of time could be as little as 20 min wouldn't want to schedule less than an hour

- 4 students at a time - each tasked with getting one full set of data

- 8 hours a day

- 16 students/day

- 5 days a week = 80 students

- 15 weeks in a semester = 1200 students (may not be using it every day during exams for example)

Key Questions

Pre-Boulder Meeting Answers So Far

Boulder Meeting Notes

Costs

What are the costs to install and operate an RWSL?

-     physical facility, power, …

-     hardware, software, network

-     staff time for setup

-     staff for operation

-     equipment maintenance & upgrade

-     consumables

Costs:

https://nanslo.pbworks.com/w/page/50971652/Costs

Equipment costs:
https://nanslo.pbworks.com/w/page/48382359/RWSL%20Sub-modules

Staffing:
https://nanslo.pbworks.com/w/page/50967107/Tech%20Staff-Support

Compare to on-campus lab?

To create a node as a minimum you need:

1. network switch and IT cabling (5K)

2. required base module (20K)

3. lab sub-module related to what lab you want to do

-- biology - microscope (20K)

-- chemistry & physics - spectrometer (15K)

55-60K would let you establish a base capability of doing biology, chemistry and physics.

If you add a half time lab manager the cost to establish a node would be approx. 100K (not including physical space).

microscope recommendation would be different if had druthers - current microscope is not reliable (China) and no local support. Would prefer Nikon (maybe corporate sponsor?) but price/microscope would be ~$40K. This also would then require labview program modification (2-3 weeks fulltime)

Need to separate initial cost vs. ongoing operational cost

Key Questions

Answers So Far

Boulder Meeting Notes

Benefits

What benefits can a “network” provide over an autonomous node?

Starting thoughts:

-          maximize capacity usage

-          extended operating hours across time zones

-          shared infrastructure and student access

-          redundancy

-          shared tech support -  FAQ, videos, manuals, …

-          shared scheduling

-          revenue and reduced costs  - quantity purchasing power/bulk buying

-          allow institutions to do what they can’t do on their own

-          relieve on-campus facility bottlenecks

-          experiment design & development

-          breadth and range of experiments and academic domains

-          shared development of expanded range of courses/credentials

-          articulation strategies and agreements

-          shared sandbox for faculty exploration

-          shared faculty training

-          consortia faculty discipline panels for continuous evaluation and improvement of open courseware and labs

-          grant applications for expanded development

-          applied research funding applications and results sharing

See: https://nanslo.pbworks.com/w/page/51638668/NANSLO%20RDN%20Services%20and%20Advantages

Need to consider downside possibilities such as loss of autonomy.

 May have to give priority to your own students first before serving needs of larger network.

Why support students outside of state or province?

Some of these benefits can be achieved even without being a consortia network

Add shared teaching to list

Use excess capacity for trading or for revenue? Or both?

Whatever capacity is unused could be available for faculty sandbox

Two levels to the network:

1. information sharing
2. service sharing

student excitement/engagement - precision of data

Key Questions

Answers So Far

Boulder Meeting Notes

Model

What is our vision for a Network?

-     centralized vs. decentralized

How do other existing consortia/networks operate and can use some of their ideas in designing our model? Eg. Internet Course Exchange (ICE), other?

Ron: I think if we want to grow this network we can't require all labs to be available at all nodes.  That just isn't practical and would drive the cost of joining too high.  I think (believe?) we are trying to build an open infrastructure that supports open education and open educational resources.  We won't get (or are less likely to get) this with a centralized model.  I envision a network that has a relatively low threshold cost to join.  Each institution will have autonomy to decide which labs they implement and how many nodes they intend to operate, but NANSLO RDN will make it economically more advantageous for an institution to serve a lab the network has none or few of, than one we have many of.

See notes at:
https://nanslo.pbworks.com/w/page/42587331/RDN%20Scaling%20and%20Sustainability%20Model

See NANSLO Consortium Membership table

Key Questions

Answers So Far

Boulder Meeting Notes

Business Model

Paul: The network needs a business model where the larger the pool of institutions and students being served the lower the price. As the network scales the cost to provide service on a per/student basis must go down. I expect the model needs to include some base level of funding coming from government along with funding provided by participating institutions. The institutional cost must be lower than what it would be if they just did it by themselves.

Ron: I agree with your general idea that cost/student should go down as the RDN grows but, while I don't mind accepting it when it is available, I am nervous about depending on government funding for ongoing operational funding.  Governments change, their priorities change, and then the funding dries up.  I do think one time government grants could be used for initial build-out when it is available and maybe for initial operations, but I hope our eventual model will not be critically dependent on government funding in the long run.  I did started thinking of a business model, but have only gotten as far as some assumptions to start with. 

See: https://nanslo.pbworks.com/w/page/51338347/Potential%20Business%20Model

See NANSLO Consortium Membership table

Tiers of membership and services – participation with and without a node?

This will require differentiation based on whether an institution is simply using services or both using and hosting services. Tiers could be tied to numbers of students or perhaps to science domains (eg. one cost if you just want biology, another cost if you want biology and chemistry, and a third if you want biology, chemistry and physics)

See NANSLO Consortium Membership table

Key Questions

Answers So Far

Boulder Meeting Notes

Tiers of Services

Can we structure participation opportunities based on tiers of services offering a low threshold to participate with opportunities for gradually increasing levels of participation and service?

Membership requirements - $, in-kind staff, other?

Ron: I'm now thinking this could be accomplished through a debit-credit system of consuming Remote Lab Services (RLS) and providing RLS rather than tie it to the number of nodes or students an institution has.  It is possible for a large institution with many nodes to serve RDN less than an institution that has fewer nodes and fewer students to use them.  That way an institution with no nodes can participate and the cost of their participation would be related to their consumption.  If they were to build a node and begin serving RLS to the RDN then the credits they receive would offset what they are paying for RLS services.  The NANSLO Board could vary the debit and credit attached to each lab so that time spent on a high demand lab would cost more than time spent on a low demand node and conversly more credit would be acquired by an institution serving a high demand or rare lab than one serving a lab that is both low demand and/or plentiful.  I think some formula would need to be derived to set the debit and credit values so this can be done in an equitable transparent way. 

See:

https://nanslo.pbworks.com/w/page/47327811/RWSL%20Distributed%20Network%20Architecture%20Discussions

See NANSLO Consortium Membership table

Considered Internet Course Exchange model where institutions list excess seats. Other institutions reserve seats (only charge if they use them). Offering institution grants reserve. Enrolling institution lists student data in database which is picked up by teaching institution and put in to LMS. Teaching institution teachers course and posts grade. Enrolling institution picks up grade and puts into SIS. Pricing model is based on $150/credit hour undergrad and $200/credit hour for grad. Also have a membership fee based on FTE.

Membership fee supports admin of consortium

supplemental instruction - students who have done well being hired as mentors for students currently enrolled - student to student mentoring online (Kansas & Laramie Wyoming)

NANSLO Consortium Membership

Consortium

Member

Dues

Phase

People

Services

Orientation

Senior admin – VP’s, Presidents, Provosts, Deans, Faculty

Demo, tour of sandbox, open courseware, site visit, research lit, orientation docs, high level funding requirements for a node, grant opportunities, institutions using it and contacts

Tier 1

$

Explore

Faculty, IT

Access sandbox work through lab exercises, not teaching students yet, supporting docs, optional training, research outcome guides (pedagogy for simulations, lab kits, RWSL), detailed funding requirements for a lab

Tier 2

$$

Pilot

Faculty & students

Option 1 use existing nodes (centralized transaction payment)

Option 2 – buy own equipment setup & use

Can do option 1 or option 2 or both

Serve own students (small number)

Tech support small scale

Tech consulting – phone for free, site support extra $

Exposure to discipline panels

Tier 3

$$$

Network User

Faculty & students

Access to tech support of students

Shared scheduling

Discipline panel participation

Governance representation (limited)

New lab request through discipline panel

Tier 4

Network Provider

Faculty & students

All of tier 3

Governance role (larger)

Provide & receive tech support

Revenue as provider

Key Questions

Answers So Far

Boulder Meeting Notes

Governance

See:

https://nanslo.pbworks.com/w/page/50888274/NANSLO%20RDN%20Management%20Model

Paul: participatory decision making on operation, expansion and enhancement of the network. May be tied to tiers of service.

Ron: Tying representation to the amount of service an institution provides to the to the NANSLO RDN was one of my original thoughts, but I've begun to shift my thinking now.  Even an institution that is only consuming RLS from the RDN needs to have some way of affecting what labs are available on the RDN and what they have to pay for RLS so they can't be entirely left out of the governance; if for no other reason than we will be dependent on the revenue they provide.  At the moment I'm thinking they could be given representation on the Discipline panels along with all other participating RDN institutions.  Since the discipline panels will make recommendations to the NANSLO board concerning what labs and how many are needed on the network this would allow them to affect these recommendations.  Institutions serving RLS to the RDN should have representation on the NANSLO Board as well as on the other parts of the governance structure.  I will update the NANSLO RDN Management Model to reflect recent changes in my thinking as indicated on the NANSLO RDN 120314 diagram I circulated today.  In an effort to get at some of this Gina tried to imagine what an annual cycle of deciding what labs will be served to the RDN.  You can see her attempt at this here: https://nanslo.pbworks.com/w/page/51837791/Table%3A%20scheduling%20RWSL%20annual%20activity   

Might be useful to have a nonprofit organization represent NANSLO to funnel donations, manage the admin and centralized scheduling.

Need to standardize things but allow growth.

Need to build international scheduling system that coordinates/pools needs and availability, (security of access & integration with LMS). CCCS building their own in house. NIC could potentially adopt and modify for own use. System needs to locally manage scheduling with excess capacity to go up centrally for access.

NANSLO needs R&D Development – development environment & developer, create new labs, multiple Albert’s linked to discipline panels, standards, conduct research           

Could have an annual/biannual Conference

NANSLO needs Management/Admin/Coordination (WICHE)

- director

- Admin

- Web site/PR

- Financial budgeting

- Grant writing & opportunities

- Membership development

- Systems development – scheduling, transactions

- Communications infrastructure – tech team, discipline panels

NANSLO Board will have subcommittees

- standards committee (technical/configuration)

- support committee

- discipline panel chairs

Boulder Meeting Notes

Fee For Service

-          Denver has capacity for non 3 pm to midnight hours – sitting idle. Have to pay for lab techs plus compensate for wear and tear, power/utilities, cost recovery. Could charge for training. For a 15 week semester to cover the lab tech time for x enrollments Rhonda has come up with $30/student. May be better to charge just on a per hour basis? Fee for student may be $60 for the kit and 10 hour access to the RWSL. Elements of the fee – lab tech time, eqpt wear and tear, consumable supplies. Scott students charged $40/student/semester. How institutions charge students for it is one thing. How we charge an institution for access is a different thing. There are two charges 1. Membership in the network (dependent on size of institution and how much they think they’ll use it) 2. Also bill out on an enorllment/use basis. 1 lab, 4 students, 2 hour lab $240?

Costs associated with usage may be best hourly. Pay for it up front to reserve the time. Could put one student or 40 students through. Approx $100/hour for use of lab.

User wants 2 microscope lab (1 hr each), 2 spectrometer labs (1 hr) (2 hr), 1 airtrack (2 hr). For section of 20 students working in groups of 4 – 5 lab groups. Need 35 hours total 10 for microscope (5 hours x 2), 15 hours for spectrometer (5 hours + 10 hours), 10 hours for airtrack. Approx. $3500

Tie membership fee to cover employees with benefits. Actual fee is for non-people time.

Could have 10 microscopes running simultaneously with the same amount of techs. 3 hours of lab time is only one hour of tech time if have 3 microscopes going simultaneously. Argues for specialization of lab facility? Costs go down the more simultaneous labs going on. Cost efficiencies are based on maximizing utilization of what is in a lab simultaneously.

Ability of lab techs to multi-task across expts is not infinite.

Makes more sense to maximize scheduling in such a way that utilizes the fewest possible hours so all three microscopes are being used simultaneously.

Approximately same cost as a lab kit.

Pilot may provide a certain amount of freebie for an amount up to x. Beyond which you pay. Tier 1 and tier 2 may be x amount per  hour. Tier 3 and 4 x amount per student.

What is the target number of institutions we need to generate operating capital from membership that lets us carve out what we need to cover staff. Then need just charges to cover non-staff.

Students pay $200 per lab kit for 11 labs – approx $20/lab

What weeks do you want to do a lab?

Part of the appeal of online is its asynchronicity. Now we are locking people in to a time. The planning and scheduling needs to be done well in advance. Need to let institutions know that we are now adding a fixed time aspect.

Build funding for development lab and sandbox into TAACCCT grant. Both development lab and sand box serve all campuses in the network.

Facility/space/IT issues related to expansion and what is required for a node. Ideal might be to double duty a traditional lab space. Would need to adapt it with the network infrastructure required. Makes it possible to have local and remote students working together. A traditional chemistry lab may be the best option? Who owns it, who is responsible for IT, if it breaks who is responsible for fixing it. Two scenarios advice on what is required for anyone creating a node (ideal down to acceptable) and finding expansion opps for our own nodes.

NGLC second round – membership expansion & support, network scheduling, multiples of equipment (focus on the high enrolment courses – biology so 3 base, 3 microscopes & 2 spectrometers, + 3 slide loaders) development lab as per TAACCCT (3base + 3 subm300K), training program (techs, faculty, developer) Be great if techs were interns leading to labour market opportunities, hire students, lab tech role for credit. Metnoring faculty, students doing reearch as tech.. What is the training/education program that skills techs, lab directors, and leads to a developer like Albert. Re-rogramming of microscope. Staff deve for 2^nd semester labs & astronomy (possibly environmental science, & engineering), computer application development (mobile, virtualization), Evaluation metrics/analytics rubric development bring in academic person for science efficacy and attitude perception/change

TAACCCT – sandbox (3 base + 3 submodules + robotic arm + slide loader $300K), expansion CCCS (3 base, 3 microsocopes, 2 spectrometer, 3 slide loader $425,000) Staff 1 CCCS FTE Lab Director $100K/year x 4 year. Lab techs (22.5/lab tech/semester x 2  = 50K x 4 year) $200K – CCCS interns, MSU- Grand Falls (2 base + 2 micorscope + 2 spectrometer), Lab tech & lab manager half time/year for 4 years. Renovate facilities, Trainng techs, Travel BC team, Interns as lab techs and job opportunities associated with lab techs (biomedical, pharmaceuticals,  …) career path for lab techs?

Pursue corporate sponsorship with National Instruments, Nikon, Ocean Optics

1

Define RWSL Distributed Network (RWSLDN)

2

Template

3

Policies

4

Procedures

5

Regional organizations of academic institutions to coordinate/contribute/maintain laboratory resources

1

Costs

2

Potential Revenue Sources

3

Faculty organization to develop and refine experiments

4

Tech staff/support contracts to operate and maintain equipment