Personal security is a key deciding factor when selecting a university or college for further education. Over 80% stated that it was ‘somewhat’ or ‘very important’ (Sem works, 2007) to their selection.Failure to provide adequate campus security can result in:
- Injury and stress
- Canceled enrolments and financial loss
- Unfavourable publicity
- Legislative breaches and fines
- Lawsuits (including punitive damages)
- Damage to reputation
- Resignation of academic staff
- Forced resignation of senior staff
- Damage and property loss
- Increased insurance premiums, or refusal of cover
Most universities and colleges respond to the cost of campus crime by investing heavily in security which includes monitored on-campus video surveillance. Unfortunately, studies show that it is impossible for a person to monitor individual feeds effectively for extended periods, (Sulman, Sanocki and Goldgof, 2008) let alone watch hundreds or even thousands of cameras. This limits traditional video surveillance to capturing forensic evidence to support prosecution. Despite significant investment in video surveillance, trend analysis does not show significant improvements in personal safety on campus compared to the broader community. Traditional approaches to video surveillance do not appear to reduce violent crime and raise privacy concerns with students from inappropriate use of cameras on campus.
The challenge is to use video surveillance to allow campus security to prevent crime while still maintaining privacy. Current best practice uses machine learning to augment a human operator by analyzing video feeds from potentially thousands of cameras and bring incidents of significance to attention in real-time. The campus security officer can then apply human judgement to assess the threat and determine the appropriate response. This is a disruptive technology that is changing the traditional campus security model. The iCetana product was developed from a university research program to support this capability. It has been deployed and used successfully at a number of large campuses around the world.
Recent threats to campus safety have reinvigorated the campus security debate and highlighted the importance of feeling safe on campus. Threats to the personal safety of students, staff and the public from security breaches are a significant risk for universities and colleges, due to reputational and financial losses. Campuses often respond by increasing their video surveillance capacity and hiring additional security staff. This document discusses campus video surveillance operations and how campus security can adopt best practices to reduce risks through improved detection and response to suspicious events.
Why is campus security so important?
A university’s brand and reputation are among its most significant assets. National and global surveys rank institutions according to their reputation for teaching and research, value for money, student satisfaction and campus safety ratings (College Stats, 2017). Obtaining high rankings involves millions of dollars spent on campaigns, slogans and sponsorship to attract top achieving students and experienced, highly-qualified academic staff. To protect their investment and maintain competitive advantage, colleges and universities take campus safety very seriously.
A survey of 12,300 prospective university students identified campus safety as a critical factor in the selection of a college or university. Over 80% of students interviewed indicated that campus safety was ‘somewhat’ or ‘very important’ to their selection. Personal safety ranked higher than residential facilities, on-campus social life, campus setting, class sizes, off-campus social life or distance from home (Sem works, 2007). Any publicised incident resulting from inadequate campus security is a multi-year threat to that institution’s reputation and economic stability.
The role of campus security is evolving. Responding to ever-increasing expectations of students and parents, today’s officers provide a human face to campus security. They act as brand ambassadors by interacting with campus community members to create security-conscious environments while simultaneously serving as the eyes and ears to create a safer place to study, work and visit.
Environment factors influencing campus security.
Substance abuse and alcohol-related violence are common factors in campus sexual assault which cost universities millions of dollars each year as a result of damages, reduced alumni donations and a drop in undergraduate applicants (Lynh and Larimer, 2017; Castellano, 2015). According to the National Institute on Alcohol Abuse and Alcoholism (2017), two-thirds of students aged 18-22 who have consumed alcohol within the past month have also engaged in ‘binge drinking’.
Given these, and other environmental factors (see Table 1), campus security faces some extraordinary challenges owing to their unique environments. Consequently, early awareness of potential warning signs is vital to reducing the likelihood of situations turning into serious incidents.
In 2015 there were 20.3 million students enrolled in universities and colleges across the United States and this number is predicted to increase to 23.3 million students by 2025 (National Center for Education Statistics Institute of Education Sciences, 2017). In Australia in 2016, there were 1.25 million students enrolled in universities (Department of Education, 2016), projected to increase by 34% by 2025 (Universities Australia, 2015). Accordingly, campus security needs to increase their presence to accommodate this influx of students.
Simultaneously, institutions continue to cut costs in the short term to bolster their longer-term sustainability due to competition from community colleges, trade schools and Massive Open Online Courses (MOOCs). This challenges colleges and universities to deliver additional value from their security expenditure by adopting best practices in video surveillance.
Campus incident trends
This analysis compares trends for criminal incidents on campus against those in wider society. Data on criminal incidents on campus across the United States was sourced from reports compiled under the Clery Act. The Clery Act mandates that any college or university that receives federal funding must record annual campus crime statistics and produce security reports (U.S. Department of Education, 2016).
Crime trends in general society were sourced from national crime statistics (the number of incidents per 100,000 persons from 2001 to 2015) (Federal Bureau of Investigation, 2016).
Figure 2 shows the trends in society versus those on college campuses for sexual offenses including indecent assault, robberies and assaults, and crimes against property such as theft. Due to a lack of statistical information and biases, the study excludes active shooter incidents, murders and arson.
RESULTS OF THE STUDY ARE AS FOLLOWS:
- SEXUAL OFFENSES - Across society, sexual offenses displayed a long-term downtrend until Alarmingly, the number of offenses reported on campus has significantly increased since 2010 after remaining constant for many years. Such increases may be due to either an increased number of offenses or an increased willingness of bystanders and victims to report offenses instead of covering them up (Lombardi and Jones, 2015). Irrespective of the cause, this is a significant point of concern for campus security.
- ROBBERIES AND ASSAULT- The trends of robberies and assaults (crimes against persons) on campus follow the trends seen in broader society, although with more variability, which is expected given the smaller sample size.
- THEFT - Theft shows a strongly declining trend relative to overall society. While motor vehicle theft has decreased over time (which may be partially attributable to improvements in car security), most of the decrease has come from reductions in burglaries. These changes may be the result of improvements in securing campus buildings through the introduction of access control and alarm
Video surveillance is an important part of a campus’ security capability which can include mobile and static guards, alarms and access control systems, emergency telephones, mobile phone notifications and other systems. According to the 2016 Campus Safety Video Surveillance Survey conducted by Campus Safety Magazine (2017), more than 9 out of 10 campuses already use video surveillance technology to reduce campus crime and 82% plan on purchasing more in the next three years.
Whilst there may be an extensive video surveillance network, with hundreds or even thousands of cameras, the information provided is often used for forensic purposes rather than helping deploy limited security resources to where they are needed. Students are often resistant to pervasive video surveillance for privacy reasons unless it is carefully managed and controlled.
A study by Bond University explains that the effectiveness of video surveillance depends on the monitoring strategies used. In isolation, the presence of video surveillance alone has little impact on crime. Only 15% of an operator’s time involved actively monitoring video feeds and searching for incidents (Wells, Allard and Wilson, 2006).
When a video surveillance operator has hundreds of individual cameras to monitor, it is unlikely that a potentially damaging event will be noticed (Sulman, Sanocki and Goldgof, 2008). Cameras become post-event, forensic tools, used to investigate an incident only when and if it is reported it to campus security. To be able to action events before they turn into serious incidents, campus security must find a way of identifying significant events in real time from thousands of routine camera feeds and actioning these to prevent crime rather than just recording it.
Best practice is not achieved by adding staff or cameras, but by adopting disruptive real-time machine learning technology, such as the iCetana solution. iCetana enables campus security to move from recording crimes that have already occurred to intervening as they occur.
An automated solution filters thousands of video feeds in real time and only highlights those that need human intervention. A video surveillance operator can then assess the situation as it occurs, then if necessary, immediately route a mobile patrol to investigate. Anomalies can be tagged and stored for later reference and assessed and actioned immediately.
Campus security is a constant concern for many college-aged students and their parents. In today’s competitive higher education market, media coverage of campus crime shapes reputation. Perceptions of campus security will either attract students or cause them to select another place to study altogether. Ineffective management of crime on campus adversely affects an institution’s reputation and imposes significant direct and indirect costs.
Whilst university and college campus have embraced video surveillance technology, they have not achieved the expected reductions in campus crimes because the effectiveness of video surveillance is limited by a person’s capacity to monitor large numbers of real-time cameras and respond accordingly. Overloading operators has resulted in video surveillance becoming a forensic recording tool for prosecution rather than prevention.
Moving to industry best practice by adopting disruptive technologies and changing security operations can address these issues. The next generation of video surveillance tools use automated real-time machine learning based on detecting and presenting anomalies for operator assessment and response. These systems provide an opportunity for a fundamental disruptive change in campus security to shift from a reactive prosecution model to a proactive intervention and prevention model.
The iCetana toolset allows institutions to drive value from their video surveillance installations and ensure that campus security can quickly address crime on campus.