Week 5 Key Insights
Key Learning Moments:
[1] “Predictive part is hard to work with because we don’t have enough data on interdictions to where we could predict anything” — CBP ReCoM Representative.
[2] “We are short sighted in looking at misses. We’re not looking at the full force laydown then and why.” — LCDR, USCG San Diego
[3] The saboteurs we are likely to encounter are the “Doubting Thomas” and “ATAK Allen” that have interests that do not align with our aims. Thomas will not accept any solution as valid until it is proven, although once convinced, he will likely be one of our strongest advocates. Allen will see us as competition to his software. Given that we do not want to set up a company after the class, we could co-opt him as an advocate by telling him what features to add to his product.
Additional Learning Moments:
[4] Patrol routes are often traced out based on human intel or intuition. Previous interdiction data appears to not play a significant role.
[5] Patrol coordinators in SD use a combination of historic data, general trends, and human intel to map out patrol routes.
- “I want 2–3 sources validating intel” (sponsor, USCG)
[6] “I want proof that this will work” (LCDR in SD)
[7] Don’t want to waste assets (chief pilot at SF Air Station, LCDR in SD)
New Hypotheses:
OLD HYPOTHESIS: patrol routes are primarily informed by historic capture data (in SD) or static pre-set zones (in LA-LB).
NEW HYPOTHESIS I: We can extract insights from past interdiction data of the LA-LB sector to increase detection, and by doing so, we will obtain buy-in from our beneficiaries.
NEW HYPOTHESIS II: Schedule visualization will be useful to patrol coordinators because they can easily plan and deconflict patrols.
MVP:
We are still not getting the most fidelity on what we should have in our product. This week, we will present two different types of MVP. One, a basic schedule coordinating mechanism to show in time and general location where assets are located across the different agencies. The second, a predictive model like a “hurricane map” to aid the agencies in allocating assets at critical locations and times.
Interviewees:
5.01 — Chief Pilot, USCG SF Air Station[Josh]
5.02 — Commander, USCG SF Air Station[Josh]
5.03 — ReCoM Coordinator from USCG San Diego [Josh]
5.04 — DHS Manager at San Diego Harbor Police[Josh]
5.05 — Graduate Student, Tambe Lab [Perry, Jacky]
5.06 — Graduate Student, Tambe Lab [Perry, Jacky]
5.07 — CBP ReCoM representative [Mike and Jacky]
5.08– Chief Warrant Officer LA/LB [Perry, Josh]
5.09 — Lieutenant at Sector San Francisco [Mike and Jacky]
5.10 — Supervisory Border Patrol Agent, CBET [Perry]
