Fall Technology Innovation Forum Day Two - Webcast

• Welcome to day two of the 2020 Fall Technology Innovation Forum. I’m Bill Alison, Berkeley CTO and your host as we showcase the Berkeley Technology Innovation Program. You can re-watch individual talks or the entire webcast from the first day at cto.berkeley.edu/innovation. We will also make today’s content available after the event as well. Berkeley is known as the home of innovation and for pushing the frontiers of knowledge.

00:26 - What we aim to do with this program is to connect those academic innovations happening all around us to the operation and administration of the university. The CTO Innovation Program has three efforts. Yesterday we showcased some of the progress being made by awardees for the first part of the program, Berkeley Changemaker Technology Innovation Grants. The grant program launched in February to fund a proof of value experiments that support our university mission and align with the spirit of Berkeley Changemaker. The second part of the program and our focus today is the Berkeley Connected Campus initiative that explores and engages the concept’s, challenges and dilemmas behind smart cities to create new experiences and efficiencies for our own campus.

01:06 - Finally, the third part of the innovation program, student impact partnerships give technology support to organizations that support student innovation, such as Berkeley Changemaker and the Blum Center we also learned about that yesterday. The Fall Technology Innovation Forum was originally planned as a day long in-person event at the MLK Center with the stress on our community as we collectively respond to the CoronaVirus pandemic, we’ve adopted Carol Crist’s charge of do less with less. We took this as a challenge to be better editors focusing on what matters the most and what makes Berkeley great. Thank you for joining us we hope you enjoy the event. Many presenters are here live in the chat to engage in dialogue and answer your questions.

01:50 - But before we get started, let’s talk for a moment about the Connected Campus. There are a growing number of smart city, as well as smart campus initiatives across the country and these can generate efficiencies, improve public safety, afford personalized experiences and improve the operation and use of physical assets all while collecting and storing actionable data about the interactions between people and things in the environment. Sustainability and preventing or mitigating environmental and health risks are among the applications we’ll see today. In IT we buy, build and maintain the infrastructure that supports these devices and services. But while the potential of smart campuses in cities and the ever expanding market for technology generates enthusiasm, it also understandably generates concern.

02:37 - Furthermore, the value of smart devices requires cross campus coordination to encourage and ensure coherent user experiences. Recognizing this, how do we promote the thoughtful orchestration of technologies to amplify human potential and increase our agency without compromising autonomy and privacy? Our nascent Berkeley Connected Campus investigates available technologies, explores their potential with campuses key infrastructure partners, parking and transportation, facilities, environmental health and safety. The Berkeley Connected Campus initiative is partnering with the privacy and security experts and policy experts at academic organizations like Citrus to develop a more scalable approach to what is already a proliferation of smart technologies being deployed today ad hoc on campus. The UC Berkeley approach also distinguishes itself from other smart campus and smart city initiatives by its people first design ethos. By this, I mean a collaborative design process that emphasizes inclusivity and empowerment for the community as a whole.

03:40 - Today’s format features talks from some people involved in building Connected Campus services, followed by a panel discussion that will dig into some of these themes. Let’s kick off a series of short presentations from our Connected Campus project leaders. Let’s start with Karen Lobo from Facilities Services, talking about our new machine learning equipment sensors. - Hello everyone, I’m Karen Lobo, director in maintenance and operations. And with me today, is Todd MaCfarlan, manager of the electrical and lightening shops. - Good morning.

04:13 - - Todd will be the implementation project manager for the smart sensor project. Thanks to Connected Campus’s grant, facility services, will install 75 smart Samsara sensors on high-value and critical equipment. These sensors function as early warning fault detection on critical equipment. And this is the gold standard in the maintenance industry. And we’re very happy that we’re moving in this direction, and thankful to Connected Campus.

04:45 - Currently, facility services operates under a reactive maintenance model. We do perform some preventive maintenance, but it’s not always driven by digital diagnostics rather the maintenance on preset schedules and visual inspections by mechanics. There can be an inherent waste in this type of model, because maintenance on equipment is performed either too frequently or infrequently. If it’s too frequent, it leads to increase labor and material cost, and unnecessary landfill waste. If it’s not frequent enough, it can lead to equipment failure and business interruption. And nobody is happy with that.

05:28 - So now I’m going to turn it over to Todd for real life example. - Thank you Karen. Currently, if a model that supports a supply or exhaust fan fails, the process is as follows: facilities is alerted up the problem, and what building the problem is at. A mechanic, responds to investigate the issue. If they find that a motor fail, they need to obtain a current quote for motor vendors. Then process it through procurement and wait for the motor to arrive, which usually takes several weeks.

06:02 - In the meantime, the building’s air system is not balanced. This introduces different potential hazards such as, security risks for doors not closing, or sometimes animals needs to be relocated. Karen will go into how our solution will change this. - So we’re partnering with Samsara sensors, they are non-invasive technology, to help improve maintenance optimization. The sensors monitor and detect vibrations on large equipment and if abnormal conditions occur, will send real-time alerts to facilities, before the equipment fails.

06:42 - This will allow facility services to be proactive, and we believe save the University money spent on unplanned maintenance and breakdowns, equipment downtime, energy, and help with utilization of personnel more effectively. Todd is going to talk a little bit about the proposed locations and equipment. - Thank you. We selected five buildings, Berkeley Way West, Chou Hall, Haas Business faculty wing, Li Ka Shing, and Northwest Animal. In most cases, we will install sensors on critical equipment or there’s no redundancy, or if failure occurs that will be a catastrophic impact to research. With pre-warning, a potential failures on equipment, we no have time to order new equipment, and schedule an outage with the department.

07:40 - We identify the critical equipment are working with shop managers, shop leads, and asset managers Cherry Chang and Karen Larson. If this test pilot is successful, will help to transform our maintenance model to a predictive one. Rather than reactor, saving thousands of dollars and overtime, and an scheduled reactor repairs. We plan to have the system fully functioning by mid February 2020. Here’s Karen. - If this test pilot project is successful as Todd mentioned it will help the campus avoid business interruption from failed equipment, lost research, and unforeseen repairs and numerous other related costs.

08:26 - The pilot project will install as I mentioned wireless sensors on assets, and will monitor and alert facilities when a component is likely to fail. Maintenance staff, will then perform the appropriate procedure to avoid the catastrophic failure. We think this is the future because we will be able to monitor the health and performance of the equipment ²⁴⁄₇ without making a physical visit. This technology enables us to establish vibration thresholds for each piece of equipment monitored and receive real-time alerts when equipment deviates. The software will allow us to have custom dashboards, perform trend analysis and set key performance indicators to give insights into bettering the efficiency and maintenance of equipment.

09:17 - We look forward to providing updates, as we’ve signed a multi-year contract with the company. And we thank you for your time. - Thank you. We will hear from Sara Souza from Environment Health & Safety talking about air sensors - Hi, my name is Sarah Souza I work with Environment Health and Safety. I’m here to provide an update on our PM2.5 air sensor project. This was started in response to wildfire smoke impacts on campus and last year we had a couple of sensors that are WiFi and plugin versions that were originally funded UCOP office of services and through the connected campus effort we were able to expand our network this year and add these solar powered and cellular transmitted devices on a new number of places on campus, both indoors and outdoors. So this is a Clarity sensor that is solar powered and transmits data via cellular networks through this antenna.

10:21 - There’s an air inlet on the bottom that pulls air in and the optical sensor counts particles of diameter 2.5 microns or less and that is what PM2.5 stands for, it’s the finer particulates in wildfire smoke that are the most harmful to people during wildfires smoke events. So one challenge we’ve had with decision-making on campus is that our reference monitors that are managed by government agencies are down at Berkeley Aquatic Park in downtown Oakland and then there’s another one further near the San Leandro border and those aren’t necessarily representative of our conditions here on campus. So by adding sensors we’ve been able to provide newer local data in comparison to those reference monitors. Currently we have these sensors installed on the roof of University Hall, the roof of VLSB, Stanley Hall, Lawrence Hall of Science and Crossroads and that really creates a representative network going up the hill.

11:29 - Since again, our reference monitor is down at Berkeley Aquatic Park at bay level that gives us much more data heading up the hill. We’ve also installed some indoors at Lawrence Hall of Science, VLSB, and University Hall, to be able to measure in real time the affectiveness of our building filtration systems and this year we were able to install some indoors for University health services as well as their early childhood education centers which are very helpful to be able to, in real time again, measure the indoor air quality and also see the affectiveness of some controls such as putting portable heavy filters inside in certain areas that don’t have HVAC systems and building filtration systems. So it’s been interesting and rewarding to work on this project this year. Our campus has been impacted for the last four years by wildfires smoke so it’s certainly a relevant issue and challenge for California. Some of the controls that we normally would of responded with are limited due to our COVID policies such as pulling in 100% outdoor air to our building, so there’s certainly some challenges this year and some pretty extreme conditions, for example, we had one week in September, many people that were around will remember the day that was orange, the skies were orange because there was smoke in the region, very high level smoke coming from the August complex in Mendocino National Forest and then our air levels on campus actually weren’t so bad that day but then when conditions changed and that smoke dropped we had some of the most extreme conditions we’ve ever recorded here on campus with an air quality index well above 200 on campus, so looking at that data in real time and understanding the impacts of some of these wildfire smokes is important for campus and for all of us that live in this community.

13:23 - It’s been helpful to see variations across campus, certainly Lawrence Hall of Science is often an outlier compared to the rest of campus because it’s higher up on the hill, and also Berkeley Aquatic Park, which is our reference monitor, because it’s down at the bay, often has very different conditions or they’re earlier, later than what was seen on campus. All that data has been valuable to understand our local trends. If you’re interested in viewing the data from these sensors you can access the data points through our campus air quality map, so it’s at the ehs website and you can click on any of the data points and then plot it in comparison to the other data points on campus or reference monitors or you’re using them out on the map you can even compare it to other regional sensors in the bay area or across California. If you have questions or interested in the data or working on similar projects, feel free to reach out to me via email. Thanks for your time. we’ll now hear a talk from Environment, Health & Safety with Bernadette. Bernadette, over to you.

14:28 - (water bubbling) - Hello everyone. My name is Bernadette Santos and with me is my colleague Tim Pine. We are with the office of Environment, Health and Safety. Today, we will be talking about the latest innovations about water quality monitoring. Tim Pine will be talking about the need and improvements in this area first and I will be talking about the Smart Cover pilot program after.

14:51 - Tim has been on the Berkeley campus for 20 plus years and has great perspectives on this topic. - Thanks Bernadette. Well, I have some good news and bad news for all of you who are watching. The good news is that incidents like this, this gushing sewer manhole happens a lot less frequently than they did 20 years ago when I was working, sorry, started working here. The bad news is when they do happen, the severity of these incidents is just as bad as it was way back then. And the key for us is early notification, the faster that EH&S can respond to an incident like this the more we can have chances to mitigate the damage.

15:31 - And the problem has been that we’ve always relied on human eyes and ears to notice the bad thing happening and then call us. And that sometimes has resulted in major delays. And this incident, this fish killed occurred as recently as last year, this was a firefighting foam release that hit Codornices Creek and ended up wiping out nearly the entire Steelhead trout population down there at Albany village. And a big factor was we didn’t hear about it and didn’t get down there until the damage was done. Also, still happening with alarming frequency, are water main breaks.

16:08 - These are because we have aging infrastructure. We have a lot of construction going on in our area. And the problem with water main breaks is the chlorine and in the drinking water that’s gushing out is equally toxic in an aquatic environment as well as all the sediment that’s being released. And then if we can get out there early enough we can actually take measures like these dechlormats these tablets actually deactivate the chlorine. And so that’s important to get there quickly to deploy those.

16:35 - And if we don’t hear about it, the worst case scenario is this damaging flow goes all the way out to the bay. And then this is the result, we have to actually put up signage all the way down to San Francisco bay. And so the good news is there’s technology now that gives us early warning of these incidents and Bernadette is going to talk about that. - Thank you, Tim. So here’s an example of our SmartCover Monitoring System. This one is installed underneath a manhole cover.

17:06 - We also have a few of these suspended over the Creek channel monitoring our stormwater system. Some of the SmartCover features include a satellite network which allows us to have better connectivity when wifi or cellular is sparse or non-existent. It’s battery powered, so the system is functional in electrical outage and batteries last about two years before needing replacement. There’s a dashboard which can be accessible by a mobile or the web. And I’ll show you a screenshot of this shortly. It has real-time monitoring.

17:42 - So the data, we capture the data every 10 minutes and it sends alerts to the users if an alarm is triggered. We have trend alerts, which allows us to see and be notified of any upward or downward flow levels, rise and fall. We have the ultrasonic sensor, which is fairly accurate. It can record data up to one 10th of an inch. And this sensor allows us to visually monitor the Creek even in the dark. It has rain monitoring.

18:17 - So we can do an overlay with the rain data and our sanitary sewer data for regulatory purposes. And it has great customer service, so if we ever need to understand the data or troubleshoot, they’re always available to help. So this is an example of the dashboard that we have via the web. As you can see, we have four sanitary sewer installations and three stormwater installations. This is an example of the chart that each system collects data.

18:54 - So as you can see on this day, we received an incident and this happened to be a foaming agent incident that went to the Creek. What had happened was when this level was triggered we received a text message to say, hey, there is a problem in this area, you should check it out. So what did we do? We had our facilities folks go to the area where we had our SmartCover and they were able to see that there was a foaming agent incident and you can see how bad it looks while on the Creek. It’s just bubbles. So this is how it’s supposed to look like in this area. So we’ve had many campus successes including two sanitary sewer overflows prevented and two illicit discharges detected including that foam agent incident we’ve been able to mitigate potential property damage and regulatory fines.

19:53 - Our engineers and planning department also has a better understanding of our sewer system and stormwater system infrastructure and facility services could also be redirected to other preventative maintenance issues and activities versus always constantly doing visual inspections at these high spot locations. So now my colleague is going to be talking about going into the future. - Thanks Bernadette. So what are we hoping to do now that we have this system? Well, first of all, we definitely want to secure more funding, so this is a soft plea for anyone who’s watching to help support this great system that we already have the beginnings of that and we’d like to do more. Secondly, the technology or the potential here is is pretty vast, right now we’re simply monitoring rise and fall in flow, but we have the opportunity to install other sensors going forward, sensors such as those that measure turbidity or conductivity. There’s even sensors available now in the market that can tell you whether there’s a sheen or rather grease floating on the surface of the water you’re monitoring too.

21:04 - And then of course we can use all the data we’re starting to gather as a planning tool and as a design tool going forward as campus, you know, begin to either replace or build out more of our sewer and storm water infrastructure. And then lastly, you know, we want to use this tool to help us engage with the campus community about the need to, you know, protect our Creek. And we still encourage people to, anytime they’re on campus, to put their eyes on the Creek. And if you do happen to see something that doesn’t look right, you, of course can always call us and there’s our EH&S number. For more information you can certainly can reach out to either Bernadette or myself, we’d be happy to talk to you in detail about this.

21:47 - And the campus also has a really excellent Creeks website. You can see the URL right there. We’ve got a page about monitoring and lots of other information about our watershed. And I want to make sure that I think, Alisha Klatt, she’s our EH&S training and instructional design specialist for helping us put this together. And of course our awesome facility services, utility plumbers. These are the folks that run out there anytime of the day or night to make sure that everything’s flowing properly and they can respond to any issues or alarms. So thanks very much for tuning in. - Thank you.

22:25 - - We kicked off the Berkeley connected campus initiative after researching useful applications of smart technologies appropriate to our campus. And examining the many potential use cases across the university, traffic management had a clear potential as a proof of value project. Part of the challenge in building a smart campus is matching useful scenarios with technology and campus units who are interested and able to participate. I’ve met with many functional teams to explore potential improvements to our campus via the connected campus and all the presentations you’ve heard today success required a willing functional sponsor, suitable technology and a real value to the unit and the university. For the technology solution we’re about to hear, Dell and NTT offered a traffic flow proof of concept pilot.

23:10 - When I propose this to Seamus at parking and transportation he was an enthusiastic partner and collaborator. He developed a proposal, one of the first for the connected campus using the Amazon Future Press Release Format I described yesterday. But after doing that he said, “Bill, this project actually makes so much sense that the unit will fund it itself if you will contribute some of your time to making sure the technology works.” I see the CTO role in part as a catalyst for digital transformation and I’m thrilled to see the fruits of this initiative. I’ll now turn it over to Seamus to describe this interesting pilot featuring observation and classification of traffic patterns for better curb management.

23:47 - - [Seamus] This is Seamus Wilmot, the director of parking and transportation, here to give you a quick update on a curb management study that we’re doing in partnership with NTT and Dell. So what we’re looking at is, the curb along Bancroft at MLK and Eshleman. And we’re going to quickly review the current state which this was pre-COVID, we’re calling it the current state, but it was pre-COVID but it was one of the busiest routes for AC Transit. We have ride hailing pickup and drop-off, we have campus and other shuttles the campus store, MLK, Eshleman and Zellerbach Hall. As a quick reminder, it’s right along here, Bancroft and the area that we’re really studying and we were able to put some NTT and Dell sensors was this intersection, the entrance to the Lower Sproul Garage that is a very busy area.

24:39 - And to give you a sense of what it looks like before COVID, so in what used to be the normal times, this is Bancroft Way looking East. This is Eshleman Hall here. And what we find is that there’s already a delivery truck in the bus only lane. We have a UPS truck that’s clogging one of the drive lanes. And the only thing we have left is this sharrow which is a bike lane and car lane. So what does this do? What does this do to traffic? Well, here we have a bus that’s going around the delivery truck.

25:12 - There’s another bus stuck behind the delivery truck and we have a pedestrian who’s just crossing in the middle of the street. Then we have just a couple of seconds later, we have cars that have to move out into the only available drive lane because the UPS truck is in the way where the bus is still stuck behind the delivery truck and we have a bicyclist weaving in and out of the cars, not very safe at all. Again, a different day but we have a truck stuck unloading its delivery right here at the bus stop. We have folks waiting at the bus stop waiting for the bus that’s never going to be able to get to them. The same BakeMark truck is halfway into the intersection that is the entrance to the Lower Sproul Garage.

25:54 - We now have a delivery truck on either side of this intersection. This one stuck at a fire hydrant, delivering its goods. This car is trying to get out of Lower Sproul Garage. We have people crossing back and forth and trucks on either side of this intersection. Again, same situation. This is one of my favorites. This truck actually broke our clearance bar but still had to get it’s delivery done so it just stopped and clogged the whole drive aisle that was getting in and out of Lower Sproul Garage.

26:26 - So how do we take a look at this? Well, we asked NTT and Dell to come in and they installed some sensors. Again, the data they’re able to do and analyze is during COVID. So it’s definitely skewed but they took a look at traffic flow and congestion. They were able to identify the types of vehicles going by, help us determine the dwell time, meaning how much time somebody is spending at the curb, what’s causing the congestion and what it just basically look like out there. So here’s one of the cameras, it’s looking East and it’s able to identify this was a bus and this picture is the second camera that’s looking southward towards Blackwell Hall.

27:04 - Again, it’s able to identify a car, identify pedestrians, another car that just went by. Here, a skateboarder would have to go behind this car that was in the crosswalk, another car going westward on Bancroft. But not only were the sensor was able to tell what type of car but they would just count to the cars for us as well. So here’s a seven day look at the cars going by in hourly chunks. And what we thought was going to be the peak time was going to be during the noon hour sort of people in the middle of the day.

27:34 - What we’re finding, and again, this is during COVID time, we’re finding that in the evenings that sort of 5 to 9:00 PM is the busiest time. So here are 7:00 PM we have 230 cars that went through in that hour. So it’s fairly interesting that in the evenings definitely is the busy time at this intersection. And what type of cars are going by? So the sensors, the NTT sensors are able to tell us that okay, the majority are standard vehicles and then about 8% are buses and public transportation which is great to see that that’s on the uptake, this is during the month of September, and then we have delivery trucks and emergency vehicles and just other types of vehicles. But what’s causing the congestion? Again, this is during COVID time so the traffic flow is much lower than normal but parallel parking causes a lot of congestion.

28:24 - People are not good at parking their cars, especially parallel parking. The bus lane, the traffic light, but here at temporary parking, people doing that double parking or pulling over to the curb and dropping somebody off that slows down a bus, that’s what’s causing the traffic congestion. So what are we going to do? What’s the future? Well, we’re hoping that these sensors, as part of our pilot that we can figure out a way to have them help us decrease congestion. Can they help us manage deliveries? Can we now use these sensors to be able to force delivery trucks to actually reserve a space on the curb? And we can tell them the time, you can only use that curb before 8:30 in the morning and you have to reserve that time. So it was only one truck delivering at a time.

29:12 - So we don’t have three or four trucks clogging up the road. Can we monetize the curb? You know, people have been traditionally driving and parking their cars but you know, the future really seems to be that you know, the ride share and you just come, you drop off at the curb and you get out of the car, you get into the car at the curb. And then also can we replicate this around to other locations around campus? At the west Press end or up by the business school or over at North Gate, other locations where we have lots of pedestrians and bicycles and buses all going through the same intersection at the same time. Can we use these sensors to really help us manage those areas? And that’s the hope and that’s why we’ll keep working with Dell and NTT to see how we can do in phase two of our pilot. - Well, good afternoon everybody. And thank you for joining our panel on the Connected Campus.

30:10 - What all the people have in common on this panel is that they understand policy and also the role of incentives in shaping desired outcomes. All too often when we talk about things like smart cities and campuses the conversation stays centered on the technology, or it focuses on a desired outcome. So the panel here has a lot of expertise in achieving outcomes while thinking about complexity and what are the second and third order effects of our technology decisions. So let me begin by making some introductions. We are joined today by Brandie Nonnecke, who is the founding director of the CITRIS Policy Lab.

30:45 - She’s also a fellow at the Harvard Carr center for Human Rights Policy. Welcome Brandie. Kira, thank you for joining us, this is Kira Stoll our UC Berkeley chief sustainability officer and our carbon solutions officer which I only learned we had when I was putting together the panel. Scott Seaborn welcome. Scott is our Berkeley campus privacy officer. And so I thought our discussion could take a different approach to discussing the internet of things and how Berkeley can take advantage of it by starting with our goals and objectives as a public research university. So Kira, let’s start with you and maybe just to begin with, can you tell us a little bit about your role and how the university thinks about sustainability? - So thank you Bill, and very happy to be joining this panel to discuss these issues.

31:40 - So sustainability has been actually part of the ethos of the Berkeley campus for a very long time. We didn’t always call it sustainability but there was moves back in the 1960s and the seventies when the environmental movement started taking traction federally and California was hit with a number of issues around drought and energy issues and oil embargoes and oil spills and so we have actually kind of a long legacy of looking at these issues from energy transportation and water, but it really took traction in the early two thousands, much in part from students that really wanted the university operations and administration to respond and reflect what we were doing in terms of our mission as a university in academics and research. And from there a really robust UC system-wide sustainability movement happened. We have a policy that has 10 sections in it, guiding what we’re supposed to do around unsustainability across UC and Berkeley in almost every case kind of exceeds those policies. So in my role as the chief sustainability officer and looking at carbon solutions for the campus, I’m focused mostly on the campus operations with drawing lines as I can to the academic and the research mission but really looking at ways that we can be resource smart and a conscientious environmental campus and provide leadership just like we do in the academics. - Thank you Kira. Scott. So let’s turn to you.

33:36 - Tell me about the role of the campus privacy officer. What do you do at Berkeley and how much of your job relates to technology and privacy? - Thanks Bill, and thanks for inviting me to participate in the panel. The campus privacy function here at UC Berkeley has a number of different roles. A lot of it has to do with compliance, so making sure that the campus complies with state and federal privacy laws, in addition to our own UC policies and privacy values. So we do everything from reviewing policies, reviewing data collection practices, monitoring practice to make sure that they’re compliant with state federal law.

34:13 - And also they’re not overbearing in terms of how much information is collected. We also work on the transparency report that Bill was instrumental in developing. It’s an initiative that involves making public every time we access personal data, personal information. In terms of this project or these kinds of projects, I think our role is really advisory. We can provide different feedback streams or provide different mechanisms for reviewing projects to make sure they’re not, and there isn’t too much data collected, we’re being transparent about what’s collected.

34:47 - If it’s shared with other entities, especially external entities, there’s some controls that are put in place. So we’re really kind of, you can say like an advisory organization that makes sure that each of these projects complies with state federal law and also our Berkeley privacy values. - Thank you, Scott. So Brandie, can you tell us a bit about citrus and your role in founding the CITRIS Policy Lab, and especially I’d like to hear how Citrus approaches navigating a balance between the potential and the risks inherent when new technologies are being adopted. - Thank you, Bill. Thank you for hosting this panel. So I am Brandie Nonnecke, I direct the CITRIS Policy Lab which is headquartered at UC Berkeley, under CITRIS and the Banatao Institute, which is actually a four-campus Institute. We’re also on the campuses of UC Davis, UC Merced and UC Santa Cruz.

35:39 - The CITRIS Policy Lab supports interdisciplinary tech policy research and engagement with the goal of supporting evidence-based policymaking, not only in the public sector, but also in the private sector especially because a lot of the governance that’s in play on emerging technologies happens within the private sector. So the question you posed about how do we balance the promise and peril of technology. I think it comes from supporting that interdisciplinary research where you connect the individuals who are thinking about the newest, greatest, most innovative technology and you’re going to partner them with people who are thinking about what are the potential benefits and the potential risks of deploying these technologies and what are the appropriate sort of leverage that we need to pull to make sure that we are maximizing the development of that technology to support benefit while mitigating those risks. So I encourage people to check out the work of the CITRIS Policy Lab at citruspolicylab.org. You can see our recent research publications on the website.

36:48 - - So Scott, how do you see policy come to life at Berkeley and sort of go from an idea to, Berkeley is not a top-down place. How, how does your role work with that? And what are the issues that you see right now with all this organic adoption of technology? - That’s a great point. It’s definitely not a top-down place. When you’re in a compliance role at Berkeley it can be a little difficult because all these different initiatives and programs develop organically across the campus and you don’t have as much visibility as you might want to. You also don’t know all the key players, sometimes even if you’re familiar with leaders and departments there’s different groups that are doing their own thing. So a lot of times it’s just building awareness of the role of the Privacy Office, what we do working with campus partners.

37:36 - We’re really lucky that there is a campus privacy group, the Campus Information Security Privacy Committee, CISPC that has members from different departments across the campus so they will keep us informed of the projects and also work directly with principal investigators, different research organizations, different campus departments to build or kind of bake in privacy practices as projects are developing. So the European Union, GDPR, I’m not sure if people are familiar with that, the General Data Protection Regulation has a concept called privacy by design, where instead of looking at a project or a data collection stream after the fact while the project’s being developed, privacy controls were put in place. I think that’s one of the things you try and do here, and we’ve been really lucky that a lot of the campus partners are already doing that or letting us know what there’s a new data collection or monitoring program, so we can provide input. - I’d actually like to piggyback on Scott’s comment about CISPC the Campus Information Security Privacy Committee. The work that they do I think is incredibly valuable and important.

38:42 - And one of the main reasons is because it promotes this diversity of viewpoint. There are individuals on that committee from all of the different departments and colleges and units, there’s faculty, there’s staff. So I think that that work is really important in any other university tuning in, I’m certain they probably have similar committees on their campus but I just want to emphasize that having these types of committees are incredibly important for helping to guide the appropriate procurement, development, implementation, and monitoring of these IOT tools on campuses. - That’s interesting, and it makes it, it raises the question for me, Brandie, of when there’s a technology evolution and there is emerging technology and the implications aren’t understood. How do those groups evolve or how do you help us have the right conversations so when a new thing, a vendor comes in and offers us something, how do we get, how do we raise awareness of what we’re stepping into? - Yeah, I think first and foremost you have to do some scenario thinking, you have to think through what are the potential pitfalls of this technology and bringing together the individuals with a diversity of viewpoints are better equipped to be able to think of those edge cases that somebody in one discipline might not think of.

40:00 - So I think really you have to do this sort of scenario thinking, they sometimes call it, this is very common in cybersecurity areas of research where you try to map out what are the greatest potential harms that are possible, kind of figure out what are the known unknowns that you can investigate. - In a real life example, Kira, when you are looking at new initiatives and there’s been a bunch that have come through your program, how I suppose you scenario planning is sort of what you do also, because on the scoping out of the benefits of new technology, you’re looking at things, how, if we’re trying to sort of think holistically about how to deal with the privacy and other issues, what as a key stakeholder leading the adoption of some of these technologies, how should we make that easy for you? When you’re going to kick off a program like BigBelly. And maybe actually you should start by telling people what BigBelly is in case they don’t know. - So I will say that the Big Bellies are, they are exterior waste disposal bins that we’ve put in place on campus. So whoever thought that garbage would be so exciting, but it was a very exciting initiative.

41:21 - And I’ll talk a little bit more about it, but I first want to kind of address the kind of the big picture when we’re looking at bringing in a new program or service, and then there’s a data component to it. And oftentimes, we’re coming at it from two points, one, we want to kind of embed this into the culture and the operations of what we’re doing. So we definitely look at how are people going to be interacting with whatever we’re bringing on board, how does it interact with the person. And then because projects the Big Bellies are so operationally focused. We also have to look through the operations lens in terms of what do we already know about something like the waste streams that we have, so using the data that we have to understand how a new tool will fit in.

42:09 - And then also being very careful when we bring new technology on, that it’s something that we can support and it’s something that we can maintain over time. And so then that also gets us into contractual thinking. Is this something that we need to manage ourselves and can learn? Or is this something that we really want to have as a service along with a new tool? So just using the Big Bellies as an example, picture a bin in Sproul Plaza at Berkeley, it’s got three containers attached with it, it’s got solar panels on the top of it, you can put organic materials, recycling and landfill in, and now you’re asking the user to make that choice and help us sort that waste. And we were kind of driven to these particular units, probably I would say the primary reason was the electronic ability or the data ability to it, but that it one, it was going to help us accomplish our zero-waste goals which are to eliminate landfill on campus by actually offering the public a way to properly dispose of their materials. And that we had this terrible system on campus of these bins that were open, critters would get in them, I mean, so we kind of also had a messy campus because of that and so we saw those solutions here with this.

43:36 - What we got and we did contemplate as we put the contracts in place. How long is the service contract for? Who owns the data? Are we going to be able to read the data? How is this, if this system gives us real-time information are we really going to be able to take advantage of it? If we know a bin is full do we really even have the staff to go out there and take care of this? And so the long story short is that the decision the campus made to move to these has had multiple benefits. It took us a while but we’ve been able to streamline some of the operations, we’ve been able to kind of reduce pickup services. So there’s been a savings and maintenance. We’ve been able to eliminate materials like bags inside, so saving money and saving resources.

44:27 - Our diversion rate is getting a little bit better, so we’re getting a little bit better sorting of the materials. And we’re also seeing a lot less trash floating around, a lot less pollution and Strawberry Creek. So, it’s overall good, how we use, what we understand from this new system, I think is still a question out there and an ongoing question for any programs we put in place as can are we really using this data that we’re gathering from these programs to our advantage. - Thank you, Kira. So what data do you collect and how much of it is used to make decisions? - We do know what our diversion rights are, so we do understand when we’re emptying those bins how much contamination there might be. So if there’s too much contamination in an organics waste stream, we have to send it to the landfill.

45:21 - What we’re finding though is that the contamination levels are particularly low so that we really haven’t seen that issue. And I think we’ve been able to, like I had mentioned before, streamline some of the operations. ‘Cause we know through the system, when the bins are full and when they’re not. The other thing that is we can reveal, these are things we knew anecdotally but it’s become more obvious with this system is that if the surrounding community doesn’t have something similar, we have a very porous border. So, materials coming in from the city, if they’re not running a similar system we’re not able to have as positive of a result, if that’s disjointed.

46:19 - - I’d like to piggyback on Kira’s point earlier in regards to who owns the data, who controls the data. If something, I think that comes up a lot and it sounds like it’s come up with this project, when you’re working with a vendor or any sort of outside organization it’s really important to put in contract terms like UC Berkeley. Whoever’s running the program at UC Berkeley in this case has control over that data. So when the program’s over, the vendor will return or destroy the data, if there’s any disclosures that are necessary to an external party that we have control over that. So I think that’s an excellent that a Kira’s team was thinking about that have been working with vendors and putting that language in contracts.

46:58 - - So it sounds almost like you’re referring to the famous or infamous Appendix DS. This data security attachment to all the procurement. And I wonder what you do when the technology works in a way that’s sort of incompatible with say the way that document was written now several years ago. Where do you get pulled in and what do those conversations look like? - Yeah, there’s been a few situations this past year when the language in Appendix DS doesn’t really work. So you have situations where Appendix DS is really written sometimes you can think of it as written for a dataset that is in one fixed location.

47:42 - There aren’t multiple users accessing it from various points. So a lot of times we will add in language in regards to putting access controls in, to systems using the data, there’s portals making sure that they’re also secure and really making sure that the vendor passes on any of their security requirements to other parties because in many cases, it’s not just the vendor that has access. They can’t even control it sometimes who has access to the data. So yeah that’s very important and sometimes we can do it and sometimes we have to escalate that process up to the campus leadership. - So thank you Scott. Brandie, so you think about this a lot in your academic work and at a very high level, I know you’ve done work with the World Economic Forum for example, advising them on their AI Procurement in a Box.

48:30 - It seems that’s very relevant to the discussion here and what do we do when a vendor says, “Sorry, Scott, we’re not accepting your terms and conditions.” We think of ourselves as a large institution, but is there a bigger opportunity for us and how can Berkeley play into that? - Yeah, thank you for bringing up procurement. That is, I think one of the strongest leverage we have on influencing how these technologies will operate on our campus. And I think unfortunately if the vendor says we can’t comply, then we’d say “Oh, well, unfortunately we can’t work with you.” The UC system if you think of it in its entirety of all 10 campuses this is a huge sector for them to work in and so if we set standards, I think that would compel them to follow in order to be able to access our campuses.

49:21 - I also want to touch on data and thinking about that in the procurement guidelines, especially when Scott was bringing up privacy by design principles, also, I’d like to talk about data minimization as a part of a privacy by design principle that ensuring these entities are only collecting the data that is necessary for them to perform the function that they’re intending to implement. And that they’re clear that only the data they need is the data they collect. And I think also there’s an important consideration about data sharing, data ownership, is data being shared with any additional third parties does the campus have ownership of that data. How does the vendor gain access to that data to improve its models and processes in those agreements that are put in place? So yeah, I think procurement is one of the most important leverage and I would recommend individuals check out the World Economic Forum’s Procurement in a Box guidelines. They’re extremely helpful, they are targeted toward government entities procuring AI technologies, but the recommendations, the guides, they even come up with questions to help individuals who are procuring the technology, questions that you should be asking vendors, they have a whole list.

50:35 - It’s extremely helpful and I think very relevant to campuses also looking to implement emerging technologies. - If I just might add to that, it’s very interesting ‘cause in the procurement area, oftentimes when I’m looking at something, I’m looking at how transparent we can be with the data or how much we can make that data available to the public so that they can interact with it as well. And so I think that that line between what’s private and being collected, the privacy piece. And then what we want to share out, to make change and transformation is really important. I’ve been on projects where we’ve been installing solar panels, they’re third-party owned and operated, but we wanted to make sure that the real time data was available to the public if they wanted to see how those panels were performing.

51:34 - And then I think the other piece I’d add to the data and the procurement piece is that also working with our vendors so that they can help us collect in a sane way, the data we need to understand how things are performing ‘cause oftentimes there’s maybe like too much data or there’s a particular one, two or three items we really need to understand and having the vendor working with us on those data pieces. - So maybe this is a question, go ahead, Brandie - Yeah, i was going to piggyback up on that. And I think you raised a really good point Kira, about maybe they’re collecting too much data, there’s like this data overload and at some point it’s just too much, you can’t gain anything meaningful from it. And I think that that is an important consideration that just because you can implement a technology or just because you can collect the data it doesn’t necessarily mean that you should. So I think this also ties back to our earlier discussion about these privacy committees, I think that those committees are really well positioned to help consider in, obviously in partnership with the entities on campus that are implementing these technologies who actually know the data that they need but to ask those tough questions about do you really need to collect that data in order for you to streamline your process to achieve X Y, and Z objectives? - And that actually raises another question and I’m not sure this may be for everybody on the group is do we collect the data is a question, and I guess another question with technology is do we even want a technology solution for something in this space? And so asking that question of things rather than then sometimes adopting the technology which maybe we’re seeking for one scenario but it actually contains a lot of other aspects within it.

53:23 - How should the university think about that when it comes to a smart campus and the idea of what we’re calling a Connected Campus because we’re, again, trying to shape our purpose around it. - I think that’s a really good question and sometimes there’s a conflict between the idea that this technology exists and that we might as well use it. And then the practical side of things maybe there’s a better way to do this the fashioned way or the old school way. Thinking about the COVID-19 situation, and some of the projects or initiatives that have been raised to keep track of who’s on campus and what buildings they’re in. Some of the proposals that come up are really exciting in terms of using wifi data, to determine who’s in campuses, buildings and things like that but sometimes just having someone campus monitor, keeping track of the numbers of people in the building in some ways it’s easier.

54:21 - So that’s just one example where the technology solution isn’t always the best solution, in this case it actually might be, but something that’s something to think about from a privacy standpoint in most cases the analog solution is just better. So sometimes you collect too much data, you may be collecting data as Brandie pointed out that you don’t need, you want to minimize it, and when you have that extra data sitting around there’s that much more chance that it’ll be shared inappropriately and disclosed inappropriately. - In the beginning part of the conversation, Brandie you talked about the CISPC committee and how it included a lot of different voices, and so we’re able to see different aspects of a problem and sort of engage in it. I’m thinking about how do we think about inclusivity as a whole in the technology choices and how they shape our physical architecture on the campus, as well as our IT architecture and when we’re purchasing things do they affect the inclusivity of our campus? - What a good question, Bill. Yes, I think so. I think all technology inherently has inclusivity barriers and opportunities.

55:28 - And so when we’re thinking about those technologies, and evaluating whether or not we should procure them, we should try to come to that decision by including the diversity of viewpoints and diversity of abilities that people have, who interact on our campus. I’ll just give one example, it’s not a really smart IOT device, but even hand soap dispensers, or like Purell dispensers, sometimes those sensors won’t recognize darker skin tones, and that’s obviously not something that we want. I think while this is one small example that can be scaled to other types of technologies that are, and you said, may disproportionately not work on certain individuals. So we have to think about whether or not these technologies are fit for purpose for all of our campus community. - In the life cycle of say, with Kira working on sustainability, how do we make sure that we inject the right conversations and considerations and make that easy, I guess you could hear from all of you, including Kira on like these are hard initiatives to do.

56:38 - It’s hard to figure out even sometimes who all the stakeholders are, what can we do have the right conversations around the technology decisions with something like this when it might not even be viewed as a technology decision? - That’s a great question and thinking about it just from the procurement lens and the sustainability aspect of it, the University of California system now has sustainability policy that says if you’re going out for a request for proposal of a certain threshold, that 15% of your evaluation criteria needs to be based on sustainability and that can be environmental, social or economic sustainability. And so already there’s this new lever that’s been built into our procurement process and I’ve just gone through an RFP process myself and it was very interesting to look at the sustainability criteria opportunity, which is so open and it could be a diverse supplier to they’ve built their headquarters and it’s a green building that they report transparently. It’s very broad right now, so there could be a real opportunity here to talk about because so much of what we’re procuring today has some sort of technological piece to it, to start defining what does inclusion mean? Even if it’s through a sustainability lens. - Yeah, just really quick drawing upon the work that the World Economic Forum has put forward I think, yeah, as Kira pointed out especially if you have projects over a certain threshold where you can implement some of these requirements. And I think that she brought up an excellent point about defining inclusion across different areas, racial, gender age, all the different types of inclusivity.

58:33 - And then thinking through, it may be in partnership with a committee like CISPC. I’m not sure if that’s the relevant committee that should do this, but coming up with a set of questions that helps individuals who are procuring the technology to ask that they’re questioning whether or not that technology is likely to be inclusive of those various categories is I think a huge help for procurement - Scott. - I think that’s a great idea and kind of piggybacking off that, the other thing that we’re trying to do more of is to solicit feedback from users who will have the experience or whose data will be collected, especially unrepresented users, who maybe haven’t had a chance to provide feedback and pass them. So I think that’s something that we’re working on, we’re trying to engage student groups at the CISPC we have two students on the CISPC, but try and do a better job of working with various student populations to make sure that their voices are heard and if there’s any sort of potential bias for example, that we can address them. - So this actually raises another question for me, which is if we’re putting a lens on the Connected Campus and sort of internet of things type applications, how should we think about establishing a toolkit for that? And how do we interleave these existing policies into something like that? Is it even the right way of drawing the boundary around thinking about something or is internet of things a component of these other policies, like you could kind of look at it from different angles and what’s in your experience in research, what’s the most effective way to approach that kind of thing.

00:16 - - I don’t have a response exactly directly to that, but one of the things to kind of point out and be able to kind of weave the connections in there there’s that we have some really urgent issues that we need to address and that I think this internet of things and this Connected Campus concept, if we could sort through some of the issues or the challenges with it’s going to be so instrumental being able to make really rapid positive shifts. I mean, we have budget constraints right now, a pandemic, a climate crisis, a biodiversity crisis. We have so many things that, and it’s so kind of difficult sometimes to get messaging out or understand how we can bring people together and get opinions that we can really use this Connected Campus idea to really help us make decisions. And it seems like there’s so many great pieces that we have right now, there just needs to be, willing to make some kind of stronger connections to help us do that more quickly. - Any final thoughts from people before we wrap up? Brandie maybe a last word.

01:34 - - I was just going to add that yeah, I think Kira’s spot on, I think that there’s this great opportunity for lesson sharing across entities on the UC Berkeley campus that are implementing these technologies, sort of the best practices where they’ve seen also where things didn’t work out. And sharing those best practices in a type of Toolkit. I’m a big proponent for Toolkit. I think that they’re very helpful, if not the end all, they’re a really good starting point. I think for organizations to be able to ramp up projects. So I just wanted to say I think that that’s an excellent idea.

02:07 - - Well thank you all and that was actually a great note to end the panel on, because this is sort of the end of the conference and our theme of innovation in the time of Corona virus. So innovation is going to happen whether or not we apply any management to it, even as difficult as the times are. So I’m glad that you all convened to help us think about how to do things so we could do less with less and still keep everything running and keep Berkeley what it is. So thank you very much. - Thank you. - Thank you. - And Thank you. - Before we end the program today, I would be remiss not to thank some of the many collaborators who have contributed to the Fall Technology Innovation Forum’s Connected Campus Webcast. We are partnered with CITRIS and the CITRIS Policy Lab, a big thank you to Executive Director Camille Crittenden, and Brandie Nonnecke, who is the Founding Director of the CITRIS Policy Lab and a fellow at Harvard’s Carr Center for Human Rights Policy.

03:05 - Thank you to Sally McGarrahan, the Associate Vice Chancellor for Facilities, and to her team, including Karen Lobo, Director of Facilities Operations, and Felix Deleon, the Director of Maintenance Operations, both of whom are involved with Connected Campus projects. Thank you to Pat Goff, the Executive Director, Environmental Health and Safety, and to the amazing EH&S team who’ve been so instrumental to our progress. A big thanks to Sara Souza, Eric Knight, Bernadette Rosero-Dugtong Santos, and Tim pine. Thank you to Seamus Wilmot, for the Executive Director of Parking and Transportation for his partnership and the great talk we just had. Thank you to UC Berkeley Chief Sustainability and Carbon Solutions Officer, Kira Stoll, for her engagement and support and joining us on the panel.

03:48 - Thank you to campus CISO, our Chief Information Security Officer, Allison Henry, Allison and her whole amazing team are great partners in keeping us all secure. Thank you to Gabriel Gonzalez, the CIO at the law school, and he’s also the chair of the Campus Information Security and Privacy Committee we referenced in the panel. Thank you to Campus Privacy Officer Scott seaborne, for all his work in collaborative thought on how to address privacy issues with the Connected Campus and everything else, Scott, thank you for joining us on the panel. I’d also like to thank our campus CIO, Associate Vice Chancellor for IT, Jenn stringer, for all of her support. David Wilson, our IT Strategic Sourcing expert, who’s instrumental when we buy these smart technologies.

04:30 - Rita Rosenthal, lead for communications and helping us get the word out on this whole forum. I’d like to thank David Greenbaum for his work on re-imagining IT, including some of the early input and feedback on the connected campus as part of Reimagining IT. The Connected Campus was an official project of Reimagining IT and the last fiscal year. Looking further afield, I’d also like to acknowledge Chuck Benson from the University of Washington for his excellent book, Managing IoT Systems for Institutions and Cities. Anyone who’s working on smart campus initiatives would be well served by acquainting themselves with his work, and also to Ben green, who I’d like to credit with influencing how we’re shaping the Connected Campus here at Berkeley.

05:10 - Ben has done excellent critical thinking and analysis on the challenges of making things smart enough. And with that, we will close the Fall Technology Innovation Forum, thank you for joining us today, and I hope to see you in the future, and you can keep tabs on what we’re doing at cto.berkeley.edu/innovation. .