Wide Area Surveillance in Support of Law Enforcement Privacy and Policy Discussions Ross T. McNutt Persistent Surveillance Systems June 2014 We request your help • PSS is concerned about privacy and the impact of these systems and has draft policies for use of its and similar systems. • PSS has been working with various police and local communities to develop appropriate policies for use in support of local law enforcement. • PSS systems are legal and covered by existing airborne surveillance laws and supreme court decisions. • These and similar technologies are coming whether PSS does it or not. • We (PSS) request your assistance to develop appropriate policies and practices that could be adopted as the standard for us and for others. – – – – Resolution Restrictions Investigation Policies Data Retention Policies Data Access Policies - Oversight Procedures - Data Protection Policies - Defense Attorney Access - Analyst Training and Agreements • We are asking for and would very much like your help Persistent Surveillance Systems • • • • • • • Sample City Unmet Needs Wide Area Airborne Surveillance Introduction Sample Coverage Areas Example Murder Investigation Analysis and Law Enforcement Use PSS and Customer Privacy Policies Request to work to develop appropriate privacy policies User Need Analysis Dayton Ohio Example - tent Surveil/ (9?5 3?onorirA?'uexnm LU -. . 4&.Jzft.? . . \layton Crime at the Neighborhood evehe" (Null-non-aw: . .2: .117. i I. . . mutt-.35: mm aux-ma . I I I. EL: mum-n; .?szn'ha'ru . .. . . I 93C um- ?lm-urqu -. - - .. a . MC: mamVL: manna; ASLLvugl PL?bE-Q?b?n?ir?ll . . . sax-memu.er ?21?9sz Lg:Nzt'. - -V I 1 an axelez?hgv \[mljj I 'u I uLz-t Fur-3m; 3.31331? - . . -- .FLIH .Inf-I Fitnur . - - - I . - I mu Isuzu . byamuse; [emu-mot. (wruzw 2; - (HM gamut-n- . . I csxs?mxanw ?l?l't . an I I k?y Dayton Crime rating from Neighborhood Scout http://www.neighborhoodscout.com/oh/dayton/crime/#data 3 out of 100 About Dayton crime rates With a crime rate of 68 per one thousand residents, Dayton has one of the highest crime rates in America compared to all communities of all sizes One's chance of becoming a victim of either violent or property crime here is one in 15. Within Ohio, more than 97% of the communities have a lower crime rate than Dayton. With a population of 142,148, Dayton has a combined rate of violent and property crime that is very high compared to other places of similar population size. Few other communities of this size have a crime rate as high as Dayton. Cost of Crime http://www.rand.org/jie/centers/quality-policing/cost-of-crime.html The cost of crime is significant. The national Institute of Justice has sponsored many studies on the cost of crime to a community. Using the results of one of these studies conducted by the Rand Center on Quality Policing the cost of crime in Dayton is calculated to be $480M per year. This is calculated by multiplying the number of crimes in a given category by the cost of crime from the study. Cost of Crime Dayton Cost of Dayton Violent Crime Type Murder Rape Robbery Aggravated Assault Number 34 93 782 602 Cost per crime Total Cost $8.649,216 $294,073,344 $217,866 $20,261,538 $67,277 $52,610,614 $87,238 $52,517,276 Total Violent Crime $419,462,772 Cost of Dayton Property Crime Burglary Larceny Auto Theft 3,390 4,882 673 Total Property Crime Total Cost of Part 1 Crimes $13,096 $2,139 $9,079 $44,395,440 $10,442,598 $6,110,167 $60,948,205 $480,410,977 Dayton Population 142,000 Cost per Person $3383 per year per person Dayton Crime Dayton has High Crime Rates 28,000 overall Crimes per year – 70-80 per day. ~10,000 Part 1 Crimes per year – 25-30 per day. 1 of 15 people per year or 1 of 4 families is a victim of crime each year 86% of property crimes in Dayton go unsolved – Typical of most cities. Impact on Community is $480,000,000 per year – $3400 per person per year. Results in lower home prices – Median $57,000 in Dayton - $158,000 in Centerville – Lower value to residence and lower tax revenue. PSS see 6 to 10 Part 1 crimes per day during operations over Dayton. PSS believes we will contribute to reducing the crime in Dayton by 20% - 30% - $96M to $144M impact. Dayton Crime Daylight Hours 1 month hr' u??emaeu - E: 332mg 35 323%; E551 is: $53 :35 E: ESE 5mm Eammagm ta :2 ea EEG Enew?? ile Square a V. 3-5' :41 cm. new] US IliZities ever Feeple Crimes Fer ?'feer Per Square Mile Cr tent Surveil/ ?93 8/70 <28 8 u??emaeu - E: 332mg 35 323%; E551 is: $53 :35 E: ESE 5mm Eammagm ta :2 ea EEG Enew?? ile Square a V. 3-5' :41 cm. new] US IliZities ever Feeple Crimes Fer ?'feer Per Square Mile Cr Baltimore 419 tent Surveil/ ?93 8/70 <28 8 Wide Area Surveillance 88 megapixels 1Hz downlinked Cincinnati 12 Sept 07 Football players practicing Wide Area Surveillance Persistent Surveillance Systems 64 Square Kilometers 192 Megapixel Color Image Operation Overview ½ Meter Resolution Recorded for analysis Electronic Zoom • Operation Overview Insert High res Ji'S'a44-min . 13:19:Will 113351 Example Investigation Murder #2 on 26 Aug 09 at 13:42 Report Number Case Number PSS Incident Locator Ingests available police reports and dispatch information Shows times and locations of flights Shows summary of reported crimes within the coverage area Searchable by location, time, crime, flight, and coverage area Significantly speeds process to identify crimes in coverage area Murders Observed and Investigated in a 2 month Period 26 Aug 13:10 - Execution of Police Woman 26 Aug 13:45 - Cartel on Cartel Murder 31 Aug 1515 - Armed Robbery 1 Sept 15:20 - Body Dumping by Liberty Bridge 9 Sept 14:09 - Attempted Execution 9 Sept 14:55 - Execution of City Official 20 Sept 13:05 - Murder 24 Sept 16:15 - Execution on Street 25 Sept 14:22 - Execution on Street 29 Sept 14:47 - Murder 1 Oct 13:47 - Execution on Street 2 Oct 17:15 – Murder 10 Oct 1407 - Murder 15 Oct 14:22 - Murder 16 Oct 14:48 - Murder 21 Oct 11:56 - Double Murder 22 Oct 15:22 - Murder in Mall Parking Lot PSS – Confidential Information Not for Public Release Many more events in data collected but we have not analyzed We witnessed many more crimes than could be analyzed with available staff Wide Area Surveillance Murder Examples Murder 2 13:47 Murder 1 13:10 88 Megapixels/sec -- up to 16 square miles coverage Ce Images of the Murder 3 Cars Directly Involved in Murder Murder Car 3 Murder Car 2 Murder Car 1 Murder Get into this car after murder Murder Location and Murderer Suspect Car 1 Path after Murder run ~11! .41" flc'et?zv' v.53)! 13 I :45 . .sv-r: Tur??nr' t? "1 '11:i?x-s "we -Iue' L- - .L :zz'mrqr'qr' nu: "r'r "s - rlcr" .543 Ham-.9 I?m 7"?Irvn 31.1! ?up: all-,1 4" 0 I-rt'. ria' .1"f1 gut". Int: anr .1 'm - 4' ?41.,I11r :1 Cooglc eartt Iools Add ?1 e226 Aug 09 arez Locations urder Car 1 Car1 prior to Carl After Murder urder Car 2 urder Car 3 urderer Path ccomplice 1 ccomplice 2 ccomplice 3 ssible Accomplice4 ar from . . I Earth Gallery a Database . rs and Labels I If 7., 1 Awareness magevy Date 5:23:2008 a 1991 Suspect Car 1 Path Prior to lVIurder I, Car1 2ndHMeetmg Starting Place on Per?on meeting Car #First MeetmgCar to Car Car1 Parking Spot 1:18:46 Suspect?hlook out location 8 minutes 1:20'35 can park'ng'spm 13'3'1'1?633r1 Wanting Location 13 4326 Car 1 4th car to car meeting 20 seconds la! 31 721567: Ion?106493534 elev 385m 31' I PTZ Cam4 14 (20(3ng eartr all 9356 Suspect Murder Car #1 Timeline 13:01:14 13:02:51 13:04:20 13:07:32 13:12:40 13:18:48 13:42:11 13:44:25 13:45:46 13:47:58 13:48:33 13:53:41 13:56:05 14:35:20 14:39:46 ……. Murder Car #1 Start Car to Car Meeting With Car 3 Car to Multi People Meeting Car to Car Meeting with Car 3 Car to Car Meeting by Where murder will happen Person exits Car to observe area where the murder will take place Car to Car meeting just down the block where the murder will happen Murder Occurs Suspect enters Car on passenger side and Suspect Car drives away 1st Person exits Car enters house Car to Person Meeting 2nd Person exits Car and enters house Suspect Car Parks - Driver exits Car Suspect Car stops (37 25.818’N 122 05.36’W) 35 minutes long Suspect Car parks and Driver exits Post Murder Locations Suspects from Car 2 Enter this House after Murder Suspects from Car 1 Enter this House after Murder Suspect Car #2 Returns here after Murder Suspect Car #1 Returns here after Murder – Drops off person Street View SUSPECT #1 STARTING AND ENDING LOCATION Murderer, Accomplices, and Associates Tracked Activities 45 minutes prior and 3 hours after murder Survey/a,7 <2 Suspect Car 1 in Systems Camera 4.12 at 13:15:27 In ?I'I'.PSS Confidential Information A A Not for Public Release Annotated Reports Integrate Ground and Arial Imagery Car 1 and Car 2 Post-Murder Activities 13:48:31: Driver from Murder Car 2 Meets Murder car 1 then enters this building. 13:47:53: Murder Car 2 returns here after murder and parks. Driver appears to exit vehicle. Corner D.M. Ruiz & Heroes del Carrizal 13:47:51: Murder Car 1 returns to location of pre-event meeting , parks briefly. Person exits passenger side (it is possible that this is the murderer) and enters this Blue building (also pictured earlier). Corner of Pascual Jaramillo & Heroes del Carrizal 24 Sept Murder Same Group I. (.ooqlc earth Robbery Spree Captured Single Perpetrator - 3 locations 5? Stat?ro?II- refuel .-.. If"? i . I .. -16mph Farrtilly Della 9 j??Starting an ma'Ending'ar'?an!?w'rxl I "the I - . I I 3.0.1.744 p" . i - 4 ans-:"a'.mmcAluv? . 23?; ?52:1; . .- . Annex Robbed ?4 7.5 .x 'r Dayton House Burglary Officers directed by PSS caught up to suspect before he got out of his car – With stolen goods in it PSS – Confidential Information Not for Public Release at Work a: Persistent Surveillance Systems Levels of Analysis • Real -Time Operational Support – Event response and information forwarding – Multiple simultaneous location surveillance – Cued or tasked support - Dispatch or 911, locations of interest, BOLO • Level 1 Analysis - Overnight – Who is directly involved with the crime – Where did they come from – Where did they go • Level 2 Analysis – 1-3 days – Who did the criminals meet with – Where did they come from – Where did they go • Level 3 Network Analysis - 1 week – Over the last week or two who did all of the above meet with – Where did they come from and where did they go Wide Area Surveillance Other Applications • Officer Support – Support for local law enforcement in crime investigations – Philadelphia, Baltimore, Dayton, Compton, Nogales, Juarez, Mexicali, Torreon, Indianapolis, Columbus, Cleveland • Major Event Security – Security and traffic management support of large events – Brickyard 400, Coca Cola 600, Sarah Palin VP announcement, 4th July Fort Leonard Wood, Ohio State Football • Emergency Response – Quick response to natural disasters and other events – Iowa Floods, Gulf Oil Spill, Hurricane Sandy • Border Surveillance – Coverage of large swaths of remote borders – Yuma Proving Grounds, Nogales Sector, El Paso Sector • Environmental Management – Pipeline surveillance, environmental impact, wild life studies, traffic Studies Iowa Floods Survey Mode Operations Imaged entire 160 miles of flooded rivers in slightly over an hour (Route shown) 70 to 100 images per location Circles over locations of particular interest 2 to 3 mile wide images See individuals, vehicles, boats, and animals Original images are 88 megapixels each Images to right are highly subsampled. Proprietary PSS Information Persistent Surveillance Systems Emergency Response Iowa Floods >1,000 sq miles surveyed in one flight Hurricane Sandy Rapid Response Imaged entire impacted areas in less then 2 hours Imagery available to emergency support Personnel 60 images per location provides moving look at any location PSS – Confidential Information Not for Public Release Privacy Policy and Concerns {egulatory lequirements PSS Privacy Policy Draft Local Policy Regulatory Requirements • Use by Law Enforcement – Reviewed by multiple city attorneys – Two Directly Related Supreme Court Decisions • Allowing Airborne Surveillance for law enforcement • Same rules as police helicopters when supporting law enforcement – Strict PSS Privacy Policy eases some concerns • Used to support police directed crime investigations • Cued from police dispatch, crime report, or ongoing investigation. – Strictly support local developed and approved privacy policies • To be included in support contracts – Political concerns remain– Careful, Caution, and Full Open Communication – Benefits are significant - Should be considered along with privacy issues • 34 murders witnessed -- confessions that account for 75 • Multiple crimes witnessed per day PSS Privacy Policies • Privacy and privacy policies are very important to PSS, our analysts, our customers, and our business • Breaches in our privacy policies can and will be dealt with as a management priority • PSS has privacy policy training and procedures for its analysts and teaches them to others • PSS resolution is limited by technology and design – Objective is 1 pixel per person and to cover as large an area as possible – You cannot tell who a person is, what they are wearing, or most the time what they are doing • Investigations are tied directly to a reported crime investigation – PSS analysts find a reported crime in the imagery then track cars and people who are involved or are likely witnesses to the crime and can assist in the investigation. – Cars and people are only tracked over public roads and sidewalks in areas that have no expectation of privacy – Occasionally during investigations – Additional suspicious acting cars catch the eye of the analyst – Cars going at excessive speeds, making multiple u-turns, not stopping at red lights or stop signs. These cars are followed to determine if they are involved in a larger crime effort. If they are not they are to be deleted. PSS Privacy Policies (Cont) • PSS analysts have had DHS suitability checks and/or DoD Security clearances to provide some assurance of personal integrity. • PSS can provide an audit trail of every location and time when an analyst looked at imagery. – PSS through its server process can record this information and managers can review this data to ensure that the analyst if looking at appropriate locations and times for the crime they are investigating. This data can show every place and time the analyst looked. Management can pull this data when needed. • PSS provides secure storage and management of the collected data. Sensitive data analyzed data can be secured in classified storage safes. • All computers are password protected and firewalled keeping nonauthorized users from the data. Previously Expressed Privacy Concerns • Police Privacy Policies need further clarification and refinement – City Council Statement April 2013 • Future cameras will be better and be able to identify people – Jay Stanley National ACLU • Looking people in their back yards / private areas - Joel Pruce – Auditable analyst viewing locations • Focusing support only on minority areas – Ellis Jacobs AL Jazzera Interview • Watching people from (Occupy Dayton) Protests – • Jobs not from Dayton community – Mr Forward NAACP Previously Expressed Privacy Concerns • Police Privacy policies need further clarification and refinement – City Council Statement April 2013 • Future cameras will be better and be able to identify people – Jay Stanley National ACLU • Looking people in their back yards / private areas - Joel Pruce – Auditable analyst viewing locations • Focusing support only on minority areas – Ellis Jacobs AL Jazzera Interview • Watching people from (Occupy Dayton) Protests – • Jobs not from Dayton community – Mr Forward NAACP Privacy Policy Development • • • • • • • • Dayton City Leadership Dayton Police Department National ACLU Local Community Groups Community Feedback International Association of Chiefs of Police Airborne Law Enforcement Association AUVSI UAV Privacy Lawyers Sample Draft Police Department WIDE AREA AERIAL SURVEILANCE POLICY STATEMENT The XXX Police Department utilizes Wide-Area Aerial Surveillance for the purpose of creating a safer environment for all those who live, work and visit the city of XXX. This policy explains the purpose of the use of this technology and provides guidelines for the operation and for the storage of captured images. DEFINITIONS Wide-Area Aerial Surveillance- Wide-Area Aerial Surveillance Sensors is an optical camera assembly that can provide continuous, second-by-second video monitoring of a city-sized area via a manned small fixed wing aircraft platform. Part 1 Crimes - The Federal Bureau of Investigation (FBI) designates certain crimes as Part I or index offenses because it considers them to be the major crimes plaguing society in the United States: Murder, Rape, Robbery, Aggravated assault, Burglary, Larceny-theft, Motor vehicle theft, and Arson. Major Events – For the purposes of this operating policy, major events will be interpreted as any event where a large gathering of individuals may congregate and there is illegal and/or dangerous activity. Weather Emergencies – Large scale damage or power outages due to weather related events. Natural Disaster Response – Coordination of emergency services to aid the populace in the event of a natural disaster. Large Scale Disturbances – Rioting, civil unrest, violent strikes, etc. Tactical Operations – Tactical operations include SWAT call outs, active shooter situations, Hostage Team call outs, maintenance of order, etc. Illegal Dumping – Large scale disposal of waste in public right–of-way. Sample Draft Police Department WIDE AREA AERIAL SURVEILANCE POLICY STATEMENT I. Purpose Wide-Area Aerial Surveillance (WAAS) can be used for general deterrence of targeted crimes; disrupting identified crime patterns; safeguarding against potential threats to the public; management of emergency response situations during natural and man-made disasters; and for support of tactical operations. II. Wide-Area Aerial Surveillance • • • • • The use of wide-area aerial surveillance (WAAS) will only be utilized to view areas of public right-of-way or public view. The current system has the ability to operate uninterrupted for five hours without having to make a stop for fuel. The sensor technology has the capability of viewing up to a 25 square mile area depending upon altitude The images captured by the sensor have the ability to be viewed nearly in real-time utilizing a down-link system. While images cannot identify specific persons, they can provide significant leads to investigators (See appendix A). • Due to the cost to operate WAAS, this technology will only be deployed for the following reasons: • • • • • • • To disrupt identified Part I crime patterns. To monitor major events, To assist during weather emergencies or natural disaster response. To monitor large scale disturbances. To support tactical operations. To monitor for illegal dumping During active WAAS operations dispatch information, to include citizen generated calls for service, will be monitored and evaluated to see if the use of this technology would assist in the apprehension or case solvability of an active incident. Sample Draft Police Department WIDE AREA AERIAL SURVEILANCE POLICY STATEMENT III. Training • Personnel involved directly with the use of WAAS technology shall be appropriately trained and supervised. • Training shall include the review of this policy. IV. Prohibited Activity • The use of WAAS will be conducted in a professional, ethical and legal manner. • WAAS will not be used to invade the privacy of individuals, to look into private areas or areas where the reasonable expectation of privacy exists. • All reasonable efforts will be taken to protect these rights. • WAAS technology shall not be used to track individuals arbitrarily or based on race, gender, ethnicity, sexual orientation, disability or other classifications protected by law. • Under no circumstances will aerial surveillance be used for the purpose of tracking persons lawfully exercising their rights to protest or assemble. Sample Draft Police Department WIDE AREA AERIAL SURVEILANCE POLICY STATEMENT V . Media Storage • • All media will be stored in a secure area with access restricted to authorized persons. Recordings not otherwise needed for criminal evidence or for official reasons shall be retained for a period of 45 days. VI . Review and Release of Video Images and related Data • • • The review or the release of video images and analytical data (to non-law enforcement personnel) shall be done only with the authorization of the Director and Chief of Police or their designee and only with a public records request. Any recordings or analytical data needed for a criminal investigation or other official reason shall be collected and stored as noted in General Order 3.02-7 and submitted into evidence as noted in General Order 1.06-1. An audit trail shall be maintained to record all access to the video images and analytical data and may be used to detect inappropriate or unauthorized use. VII . Policy Violations • • Unauthorized access to the WAAS recorded imagery, misuse of the imagery, unauthorized reproduction of the imagery, or unauthorized distribution of imagery will result in an Administrative Investigation. The Director and Chief of Police or his designee shall authorize access to the system. Previously Expressed Privacy Concerns • Police Privacy policies need further clarification and refinement – City Council Statement April 2013 • Future cameras will be better and be able to identify people – Jay Stanley National ACLU • Looking people in their back yards / private areas - Joel Pruce – Auditable analyst viewing locations • Focusing support only on minority areas – Ellis Jacobs AL Jazzera Interview • Watching people from (Occupy Dayton) Protests – • Jobs not from Dayton community – Mr Forward NAACP Resolution • 1 Pixel per person is what is desired and as large coverage area as possible • Current Resolution ~1.5 - 3 ft per pixel ~ 1 pixel per person – 4 pixels per person provides no additional useful information – 9 pixels per person provides no additional useful information • Resolution needed to – Identify persons color of clothes – 20-25 pixels per person – Identify persons race – 50-75 pixels per person – Identify person 300-1000 pixels on face 1 sq ft – 1000 times improvement • Would need > 1,000 improvement – Looking at top of head. Our Resolution Objectives • Be able to track humans from crime scenes to vehicles • Be able to track vehicles from crime scenes to place of origin and place of destination • Maximize coverage area PSS Proprietary Information Coverage Area 25 Sq Miles 5 4 16 Sq Miles 3 9 Sq Miles 2 1 4 Sq Miles 1 Sq Mile PSS Proprietary Infomration . .54? n.r. 27 Feb 1 IOHS I a . .. .. .. zau.m.\? . . gnaw.? . .. .. .. . ..Kw..mw ?mutamMurder of Israel Munoz Oronday Murder Shot at 13:40:42 - 27 Feb 2010 Victim Car Accomplice Shooter running toward victim car Murder Car FOUO - For Official Use Only The Murder of Luz Adriana Cobayachi Peñaloza & Aldo Alan Contreras Cobayachi Murder Shot at 13:56:16 – 27 Feb 10 Shot of actual shooter, victim and getaway car Accomplice car 2 (left) also shown waiting around the corner PSS Proprietary Infomration Our Resolution Objectives Are Not To • Identify vehicles specific type of vehicle – 6 inch resolution • Read License Plates – <3 cm resolution - wrong look angle • Identify people – <1 cm resolution – Wrong look angle – top of head • Determine if someone is armed – 2 cm resolution • If I had more pixels would want to cover more area and see more crimes, track more criminal to final location PSS Proprietary Information Previously Expressed Privacy Concerns • Police Privacy policies need further clarification and refinement – City Council Statement April 2013 • Future cameras will be better and be able to identify people – Jay Stanley National ACLU • Looking people in their back yards / private areas - Joel Pruce – Auditable analyst viewing locations • Focusing support only on minority areas – Ellis Jacobs AL Jazzera Interview • Watching people from (Occupy Dayton) protests – • Jobs not from Dayton community – Mr Forward NAACP PSS Analyst Training • 80 hours of hands on analysis training • Training Formalized as Community College Courses • • • • • • • • • • System Introduction Privacy Policies and Procedures System Uses and Limitations Tracking Investigation Process Intel process Criminal TTPs (What they typically do) Integration of sources Reporting results Leading teams Reviewable Viewed Areas and Tracks • Analysts support investigation of reported crimes – Analysts track people and vehicles in the proximity of reported crimes. (Potential suspects and witnesses) – Analysts are not allowed to view areas or track vehicles that are not in support of investigations – Tracks are identified by in investigation supported and person tracking • Recorded viewed location by analysts – Images are served from central servers by requested location and time – Lat, Lon, time, and time of requested images from server by analyst • Viewed areas are reviewable by management and others – Allows verification of areas looked at, time, and by whom – Reviewable by others to verify supported areas Sample Server Log User Time Of User Request RossMcNutt RossMcNutt RossMcNutt RossMcNutt RossMcNutt RossMcNutt RossMcNutt RossMcNutt RossMcNutt RossMcNutt RossMcNutt RossMcNutt RossMcNutt RossMcNutt RossMcNutt RossMcNutt RossMcNutt RossMcNutt RossMcNutt RossMcNutt RossMcNutt RossMcNutt RossMcNutt RossMcNutt RossMcNutt RossMcNutt 05/13/14 13:33:25 05/13/14 13:33:25 05/13/14 13:33:25 05/13/14 13:33:27 05/13/14 13:33:27 05/13/14 13:33:28 05/13/14 13:33:28 05/13/14 13:33:28 05/13/14 13:33:30 05/13/14 13:33:32 05/13/14 13:33:32 05/13/14 13:33:33 05/13/14 13:33:33 05/13/14 13:33:34 05/13/14 13:33:34 05/13/14 13:33:35 05/13/14 13:33:36 05/13/14 13:33:37 05/13/14 13:33:37 05/13/14 13:33:38 05/13/14 13:33:38 05/13/14 13:33:39 05/13/14 13:33:39 05/13/14 13:33:40 05/13/14 13:33:40 05/13/14 13:33:40 NorthLatit WestLongit SouthLatit EastLongit FrameNum ude ude ude ude ber Time of Data 39.7754 39.77964 39.78418 39.7851 39.77968 39.77548 39.77283 39.77028 39.76855 39.77045 39.77261 39.77507 39.77512 39.77143 39.77143 39.76812 39.76812 39.76518 39.76518 39.76285 39.76285 39.76217 39.76213 39.76052 39.75884 39.75771 -84.2072 -84.2146 -84.2228 -84.2157 -84.2077 -84.2017 -84.1979 -84.1942 -84.1917 -84.1945 -84.1978 -84.1934 -84.1932 -84.1854 -84.1854 -84.1817 -84.1817 -84.1831 -84.1831 -84.1898 -84.1898 -84.1934 -84.1934 -84.1922 -84.191 -84.1901 39.745882 39.740798 39.735035 39.735948 39.742328 39.747103 39.750093 39.752974 39.754868 39.756773 39.758931 39.76139 39.761442 39.757756 39.757756 39.75444 39.75444 39.751504 39.751504 39.749172 39.749172 39.74849 39.74845 39.749122 39.74983 39.750305 -84.166663 -84.161232 -84.155281 -84.148211 -84.156426 -84.162698 -84.166637 -84.170426 -84.172956 -84.175678 -84.17901 -84.174567 -84.174432 -84.166585 -84.166568 -84.162888 -84.162871 -84.164316 -84.164299 -84.171044 -84.17101 -84.174599 -84.174599 -84.176585 -84.178609 -84.179949 464830 464830 464830 464830 464830 464830 464830 464830 464830 464831 464841 465011 465021 465031 465051 465061 465071 465081 465091 465101 465111 465121 465131 465141 465151 465161 03/31/14 16:54:43 03/31/14 16:54:43 03/31/14 16:54:43 03/31/14 16:54:43 03/31/14 16:54:43 03/31/14 16:54:43 03/31/14 16:54:43 03/31/14 16:54:43 03/31/14 16:54:43 03/31/14 16:54:44 03/31/14 16:55:01 03/31/14 16:55:02 03/31/14 16:55:03 03/31/14 16:55:05 03/31/14 16:55:06 03/31/14 16:55:07 03/31/14 16:55:08 03/31/14 16:55:09 03/31/14 16:55:10 03/31/14 16:55:11 03/31/14 16:55:12 03/31/14 16:55:13 03/31/14 16:55:14 03/31/14 16:55:15 03/31/14 16:55:16 03/31/14 16:55:17 Previously Expressed Privacy Concerns • Privacy policies need further clarification and refinement – City Council Statement April 2013 • Future cameras will be better and be able to identify people – Jay Stanley National ACLU • Looking people in their back yards / private areas - Joel Pruce – Auditable analyst viewing locations • Focusing support only on minority areas – Ellis Jacobs AL Jazzera Interview • Watching people from (Occupy Dayton) protests – • Jobs not from Dayton community – Mr Forward NAACP Covered Areas • Up to 25 square miles – Covers multiple neighborhoods and communities at once – Systems does not focus on single areas • Where/when to support determined by Law Enforcement – Based on reported crimes, statistics, and crime heat maps from Dayton Police information system – Locations, times, and severity of crime patterns • Working from Reported Crimes – Victims/Citizens have requested support by calling 911 – If they do not call we do not know about it and can not support - tent Surveil/ (9?5 3?onorirA?'uexnm LU -. . 4&.Jzft.? . . \l . I . 'r "fir,IlJukf?ngy3.?le ?l .1 6 . ?ll-"71?? ?v Previously Expressed Privacy Concerns • Privacy policies need further clarification and refinement – City Council Statement April 2013 • Future cameras will be better and be able to identify people – Jay Stanley National ACLU • Looking people in their back yards / private areas - Joel Pruce – Auditable analyst viewing locations • Focusing support only on minority areas – Ellis Jacobs AL Jazzera Interview • Watching people from (Occupy Dayton) protests – • Jobs not from Dayton community – Mr Forward NAACP Summary • New and Unique technology • Demonstrated significant impact on law enforcement operations • Success in investigation support across a wide array of crimes – 34 murders witnessed and investigated – Confessions that account for 75 – Kidnappings, Robberies, Assaults, burglaries, car theft, car jackings, … • Multiple other uses – Event Security, Emergency Response, Border Security • Affordable system aimed at large to mid-level police departments – Purchase, lease and service options available • Working to ensure appropriate policies to ensure protection of people and privacy – Asking for your input and comments nee Good People Doing Good Things Wide Area Surveillance Other Applications • Officer Support – Support for local law enforcement in crime investigations – Philadelphia, Baltimore, Dayton, Compton, Nogales, Juarez, Mexicali, Torreon, Indianapolis, Columbus, Cleveland • Major Event Security – Security and traffic management support of large events – Brickyard 400, Coca Cola 600, Sarah Palin VP announcement, 4th July Fort Leonard Wood, Ohio State Football • Emergency Response – Quick response to natural disasters and other events – Iowa Floods, Gulf Oil Spill, Hurricane Sandy • Border Surveillance – Coverage of large swaths of remote borders – Yuma Proving Grounds, Nogales Sector, El Paso Sector • Environmental Management – Pipeline surveillance, environmental impact, wild life studies, traffic studies Persistent Surveillance Systems • Conducting Live Operations since 2008 – – – – – – – – – – – – – – – – – – NASCAR BrickYard 400 Indianapolis Motor Speedway North Philadelphia Philadelphia Police Department Baltimore Police Department NASCAR Race Cola Cola 600 Charlotte Motor Speedway Iowa Floods Dayton Police Department Sarah Palin VP Candidate Announcement Ft Leonard Wood Juarez Mexico Private Kidnapping El Paso Border Surveillance DEA Gulf Oil Spill Mexicali Mexico Torreon Mexico Compton California – LA Sheriffs Office Nogales Arizona Border Security Operations Dayton Police Department Hurricane Sandy Response NJ/NY Emergency Response Major Event Security • • • • • Pre-Race Planning Minor Incident Response Major Incident Response Traffic Management Business Case Brickyard 400 Indianapolis Motor Speedway Emergency Support and Disaster Response Real-Time Major Incident Response: • Fast wide-area damage assessment • Coordinate mass-evacuation, perimeter cordon • Common Operating Picture for Police, FBI, emergency responders, and other event personnel • Detect movement over broad areas for potential survivors Proprietary PSS Information Rewind imagery to track and backtrack potential suspects and witnesses Staring and Survey Modes Persistent Stare and Long Range Coverage Persistent Coverage over and Area -- Used for high activity areas -- Constant Coverage over large area -- Detect and Track Movement Within Area -- Real Time Constant Download Long Range Periodic Coverage -- Used for Long areas of interest -- Multiple real time images allow for image to image change detection to auto detect movement within the area -- Pass by pass detection can highlight changes that occur between over flights -- Periodic Downlink when within range Iowa Floods Survey Mode Operations Imaged entire 160 miles of flooded rivers in slightly over an hour (Route shown) 70 to 100 images per location Circles over locations of particular interest 2 to 3 mile wide images See individuals, vehicles, boats, and animals Original images are 88 megapixels each Images to right are highly subsampled. Proprietary PSS Information Persistent Surveillance Systems Emergency Response Iowa Floods >1,000 sq miles surveyed in one flight Hurricane Sandy Rapid Response Imaged entire impacted areas in less then 2 hours Imagery available to emergency support Personnel 60 images per location provides moving look at any location PSS – Confidential Information Not for Public Release