General Electric

Team: Sanket Shukl, Maggie Criqui, Alex Brown

Responsibilities:  User research, Sketching, Ideation, Journey Maps

Duration: 1 week

Type: Graduate course project under Prof. Marty Siegel

Hypothetically, imagine that IU and GE have installed thousands of sensors all over campus to capture everyday life. These data can be whatever you want it to be –weather data, cell phone or car geolocation information,facial recognition, crowd estimation, to name a few. Push the boundaries of your imagination about what information can be tracked, and how it would be useful to solve a campus problem (such as providing traffic or parking information during finals week).

Your challenge is to select a student or campus problem and then design a solution that addresses it,based on:

• The data you will use to solve the problem .
• The outcome you are trying to achieve.
• The type of analytics you would use to achieve this outcome

Step 1: Understanding Industrial Internet of Things (IIoT)

The key to understanding what our client (GE) really needs was to understand the Industrial Internet of Things. GE has a lot of services offered by Predix, that follow this concept and leverage the interconnected sensors and use their data to predict wear and tear, usage statistics, logistics etc.

Step 2: Understanding systems design

To understand and appreciate the interconnected nature of parts in how a system works, gave me and our team a baseline of questions and constraints that we need to have in place. I referred a book called 'Thinking in systems- a Primer' to enhance my knowledge of feedback loops and how to identify the interconnections in a system.

Step 3: Tying step 1 and 2 together

Once we had set our own definitation of IIoT and what exactly a systems desgn entails, it was easy for us as a team to brainstrom about possible problems we see on campus and which ones we would like to solve. We categorised the problems into broader topics and then decided to narrow down on one topic.

After brainstorming about possible problem spaces, we noticed that the biggest problem on campus that could benefit from GE's IIoT software Predix was the problem that plagues every building on campus. The HVAC system in every building is different and students have no control over the inside temperature and humidity.

Often it so happens that a class room is too hot or too humid and students ends up opening the windows or doors to compensate for the lack of temperature control. We observed, specially in dorms, this problem is more severe and annoying since during winters the rooms get more hot than expected due to HVAC malfunction and students end up opening the windows. This leads to electricity wastage and further deterioration of the HVAC systems in place.

Since we had the liberty to install sensors and collect data that we thought was useful to us, we decided to solve the problem of HVAC and make all the buildings on campus 'SMART' by automating temperature control , irrespective of the types of HVAC systems used.

We first noticed that students and teachers were leaving their windows and external doors open even if it was very cold or hot outside.

When asked, they said that the rooms were always an uncomfortable temperature and facilities could not seem to resolve the issue.

So we reached out to the facilities team to get to the root cause of the issue.

Our next step was to get in touch with the facilities manager for our department building and ask him a lot of questions about the working of the HVAC system. The interview with the facilities manager was really fruitful as we got a lot of insights and he even showed us how he works.

Insights:

• Every building on campus has a different HVAC system mechanism in place

• Some systems are archaic and need lot of monitoring

• There are two main types of HVAC systems- Pneumatic and Polyline

• Certain systems take longer than others to adjust the room temperatures

• Facilities manager does not have all the access in terms of controlling the old systems

• He has to manually set the thermostats and keep monitoring them

From our understanding of the interdependent parts of the entire HVAC system, we identified two way key aspects for our proposed solution:

1. Automation

With the use of sensors, the repetitive tasks under the 'routine' category can be automated for better precision and avoiding laborious tasks for the facilities manager. The time saved with this solution in place can be spent by the facilities manager on attending to issues that need his attention.

2. Proactive Monitoring

The sensors can also be used to continuously monitor the temperature and provide accurate readings. With a proactive alerts monitoring system, the facilities manager can know well before hand if any particular room or office needs attention and he can proactively fix the issue before the situations gets worse.