Startup Policing: Lessons learned from having a cop work in a university startup incubator

Written by:

Shawna Coxon, deputy chief of police with the Toronto Police Service
AND
Ian Williams, manager of analytics and innovation with the Toronto Police Service

To create new technological opportunities, police departments must develop new partnerships in unconventional ways. This article explores one opportunity used by the Toronto Police Service in Canada to leverage unorthodox thinking to find unique partners willing to help solve persistent public safety challenges.  

Background 

Toronto, Canada is the 4th largest city in North America and is growing at a rate three times higher than any other city in the U.S. or Canada.i If the growth rate continues, the city is posed to hit 3 million people by 2022. One of the reasons for this spectacular growth is the booming tech sector, which is transforming the city.

In 2019, the Economist Safe City Index ranked Toronto the safest city in North America and the 6th safest in the world. This is a key reason the Toronto Police Service (TPS) employs far fewer members than comparable cities internationally; there are approximately 4,800 officers and 2,200 civilians in the Service. Just as the city is transforming to a technology-driven community, TPS undertook an enterprise-wide modernization initiative, which significantly affects every area of the police organization, including technological innovations.  

The story of rapid transformational change, including rising demands by the public along with criminogenic problems that are increasingly complex, is a situation facing many police agencies internationally. Managing these challenges require innovative partnerships that can assist in researching, assessing, tackling or ‘hacking’ public safety problems using novel perspectives. The Toronto Police Service recognized it was sitting in the middle of a rapidly booming technology sector and looked for partnerships that could assist in driving police innovation. They looked to a local university, which runs the #1 ranked university-based start-up incubator in the world.iv .

A start-up incubator is a program to help new companies grow their business. Such programs provide workspace, mentoring, training, and sometimes seed money to help the start-up get up and running. More and more universities are offering such programs in order to attract students who want their education to include starting a real business. University incubators have been found to be highly successful, leading to more jobs and higher sales than private sector incubators.v Ivy league schools such as Harvard and the University of Pennsylvania offer start-up incubator programs, as does Berkeley, MIT and Stanford.vi 

In Toronto, Ryerson University lies in the heart of the city and has over 45,000 students enrolled in over 100 undergraduate programs. The university is renowned for its tech start-up programs, including its world-class start-up incubator called the DMZ.vi With offices in both Toronto and New York, start-ups embed within the university-based program are connected with capital, customers, and methodologies from a community of experts, entrepreneurs, and influencers. The goal of the program is to support new technology businesses until they scale out of the start-up phase.ix While this model is common and accepted for the private sector, having a direct partnership with start-ups in the government sector is challenging due to government procurement rules and local political concerns of bias in police getting too close to a particular vendor. Yet, it is common for police to work collaboratively with universities on community safety challenges. The Toronto Police Service met with the Executive Director of the Ryerson University DMZ to discuss working in partnership in a different way.  

Why embed a cop into a university start-up incubator? 

Members from the Toronto Police Service and the DMZ met and decided to embark on a unique police/university partnership. TPS agreed to embed a frontline police officer with vast field experience into the university start-up incubator for a minimum of 6 months. The university provided working space and access to the same communities of experts that private start-ups receive. Such a program had never been done before in Canada or the U.S.; no public sector agency had partnered with an incubator in this way.  

At the outset, the outcomes of what to expect at the end of six months were also somewhat vague, which was difficult to explain to internal and external critics of this initiative. What was evident from the beginning was that this new process would allow for completely different thinking, an unusual methodology to consider persistent or emerging policing problems, and this new strategy created great potential for better policing outcomes. It was a pilot project in looking at the kinds of questions and assessments traditionally asked about policing challenges to see if other questions could be posed in order to increase the probability of solving identified problems.  

Given that this program was completely novel, the officer was chosen carefully. He was a frontline police officer working in the busiest police station in Toronto, however, he had repeatedly emerged as someone who wanted to challenge the status quo. He was known to come up with creative solutions to frustrations faced by frontline officers, often to the exasperation of his supervisors. His perspective meant that he arrived at the university start-up incubator with a list of tangible challenges facing officers, particularly around inefficient processes, and this project gave him an opportunity to actively work toward researching potential solutions.  

Another key component was the steering committee created for this project. This committee included a variety of stakeholders interested in advancing technological innovation in policing. The Toronto Police Service representation on this committee included a Deputy Chief and the Manager of Analytics and Innovation, as well as the Executive Director of the Ryerson University DMZ. Rounding out the steering committee was the director of the City of Toronto’s new Transformation Office, as well as high-level participants from various private sector companies. This cross-sectional approach with experts in innovation from a wide variety of backgrounds was critical to the project. Everyone on the steering committee was highly engaged and had a genuine interest in fresh, outcome-based solutions. The steering committee met every two to three months and their feedback was critical to this project’s success.  

What can police agencies learn from start-ups?  

Over the next six months, key lessons were learned and quickly applied within the Toronto Police Service. While this project began as an investment in questions where new methodologies would be used to consider problems, the lessons learned quickly led to tangible outcomes. Below are the top 4 lessons that were rapidly adopted. Each one will be discussed, including how each added value in moving specific initiatives forward.  

• Agile Methodology  

• User Design Expertise  

• Progressive Procurement  

• Innovative Research through the First Policing-Led Hackathon in Canada 

Lesson 1: Agile methodology 

Project management in policing has typically followed a waterfall methodology,x where the entire project is mapped out in detail at the beginning and this plan must then be followed one step at a time without ever going back. The entire project is then delivered all at once at the end. This methodology is highly bureaucratic and makes sense for large, enterprise-wide technology projects.xi For example, this approach would be suitable for implementing a new records management system. But smaller projects could benefit from a faster, more nimble approach. Contrary to what was happening in TPS, none of the start-ups in the incubator were using a waterfall methodology to develop their products. Instead, they were leveraging agile methodology, which is an iterative approach that delivers products in increments through collaboration and the ability to jump back and forth through project steps.xii This agile approach is more beneficial for software development, particularly for products that are tailored to the end-user, such as apps and dashboards. TPS found that internally all technology projects were being developed using a waterfall style of approach and agile methodologies had not been applied. In some projects, this was found to be cumbersome and led to frustration.  

TPS began using agile methodology on smaller projects, but the newness of using the approach ran counter-cultural to the thought processes used in many policing organizations. In TPS, people didn’t understand the model and believed it would compromise projects. The importance of failing fast or using iterative models to deliver sections of the product in short timeframes all made people uncomfortable because these were things people were traditionally penalized for. The Manager of Analytics and Innovation quickly realized the benefit of such an approach for information projects such as apps and dashboards. He took the training to become a Certified Agile Practitioner and created an Analytics Centre of Excellence (ANCOE). This team received training in agile methodology, and this eventually led to further training in other sections of TPS. Now, having been adopted as an effective methodology, both traditional waterfall and agile methodologies are used based on the project needs. This has led to more effective, faster development of user apps and dashboards that officers and the public can use to access the public safety information they need.  

Lesson 2: User design expertise 

While projects have always considered what outcomes end users desired, the start-ups in the incubator heavily leveraged user design (UX) experts. This is a fairly new area of expertise not often used in the public sector. A UX designer looks at how products are used and understood by end-users. For example, they will look at how the product is designed, used, how it functions, and what story/branding needs to be put around the product for the end-user. All of this is considered and implemented before the product is rolled out to ensure products will be valued and used by those they are meant to serve. This kind of expertise adds value to both members of the police service and the public, depending on what product is being rolled out.  

One key game-changer that came out of working with UX designers was the development of detailed personas for both members of TPS and the public. Personas are a tool commonly used by UX designers and product developers. The idea behind personas is to create the identity of a typical user. The user is not a real person, but a representation, and it is meant to be very specific in terms of what that user would want. For example, several personas would be created regarding different officers who work out of a particular police station or of citizens who live in a particular neighbourhood. These personas are based on detailed research usually consisting of interviews and surveys. They are then used when products are being developed to ensure the needs of the personas are being met as the technology is being developed. The personas help the developing teams to continuously focus on the end-user.  Personas are only one tool, as it is still important to find out the perspectives of actual users of a product as it is developed.  

Related:

Challenge-based RFPs and open data keys to solving Toronto’s biggest problems, municipal leaders say 

 

Understanding the value of UX expertise, a UX designer has been permanently hired by TPS and TPS members were also trained on how to run UX workshops. Surprisingly, this has been extremely beneficial not only for internal projects but also for community engagement. While officers can map out how they would like various products to work, community members have been able to map out what they feel is important to achieve safety in their community. This approach allows all members present to have a voice, as well as members who aren’t able to attend in person via UX surveys.  Neighbourhood officers have been able to prioritize projects, build teams within the community with stakeholders not previously engaged, and collaboratively map out desired end states using UX methodology. 

Lesson 3: Progressive procurement

This partnership worked because TPS was working with start-ups through the university incubator. Since the partnership was with the university and not with any specific private sector company, police were able to work alongside start-ups and learn from their problem-solving approaches. There remains a challenge in collaborating directly with private sector companies to co-create a product because the company would then potentially be precluded from any future procurement process. During the project, this became an increasing frustration point because TPS was losing the opportunity to access new and emerging technologies.  

A typical procurement process in Canada would start with a Request for Information where general research is conducted regarding vendors in the market and what their products do. This would then lead to a Request for Proposals (RFP) in which the technical specifications needed for the project are published and then vendors use these specifications to compete for the contract. The problem is when a new and emerging technology needs to be developed specifically for the organization and this means that the technical specifications are not known until the vendor is working with the police agency to understand their systems. This kind of co-creation is not permitted until the contract is won and the vendor is brought onboard. But the RFP process presumes that such specifications will be known by the police agency and used in the competitive process. In other words, the police need to work with a vendor to design a product including the technological specifications, but the vendor can’t do so until they’ve been hired, and they can’t be hired until these specifications are known. This becomes a circular argument with no procurement resolution. Further, the cumbersome nature of this process prohibits many start-ups with cutting edge technology from competing. This is a problem acknowledged at the highest level of government in Canada who is also looking for more progressive procurement solutions.

This challenge has led to a desire across the public sector to work in an accountable and transparent manner with vendors to co-create for product development.xiv This same issue has led some agencies in Europe to move toward paying for services based on outcomes rather than detailed technical specifications to drive public sector and vendor collaboration.xv This is particularly important to be able to leverage emerging technologies where organizations may not yet understand such specifications. A common refrain is: I need to know what I want before I can go to market, but what if it still needs to be developed? What if it’s not clear to me what I need, and I only understand the problem to be solved? How do I procure for that?

To tackle this problem, TPS decided to trial a partnership procurement initiative which was organized around publishing a problem to be solved, rather than specific technical specifications. To do so, TPS issued an RFP that was structured as an Invitation to Partner (ITP). The ITP procurement model had only been done once before in the City of Toronto through their Transformation Office, but it had not been used by any other city agency. The ITP was for a technology project around the use of automation in the 911 call center. It called for a straightforward disclosure process to be automated and if this project was successful, a 3-year partnership could be signed. This would mean that the company hired could work collaboratively with TPS to determine and then design future processes for automation, including co-designing the technical specifications which can take months to determine, and then implement them without going to market again. This was a very challenging process to implement, but the first disclosure automation is now built, the 3-year contract is signed, and the next process for automation is currently being c0-designed by TPS and the vendor. The ITP procurement model was essential to allow for police-vendor technology collaboration on this project.  

While the ITP model allows for co-creation via a partnership, TPS is now researching increasing collaboration even further by mapping out an outcomes-based process which would allow public-sector organizations to run design competitions and then procure the winning product. Such procurement competitions are already used in Europe, where events such as hackathons are run to create and implement new, innovative technologies. Although public sector agencies in Canada have not figured out how to run such competitions as a procurement process yet, there is still value in the hackathon model for police departments looking to research emerging technologies.  

Lesson 4: Innovative research through the first policing-led hackathon in Canada 

Wait – what is a hackathon?  

A hackathon is a timed, sprint-like design event that often takes place over 48-72 hours, although it can take place over a much longer period of time depending on how complex the problem is.  The goal is to create a usable product that addresses a problem statement identified by the sponsor at the beginning of the hackathon. Through intensive team collaboration and design engineering, teams are able to present new and innovative solutions that address the problem statement.  Ideas created during these events are often unique and creative.  Teams are comprised of subject matter experts. For example, a technology–based hackathon would attract teams made up of project managers, computer programmers, machine learning experts, graphic and interface designers and other related experts.  Hackathons have often been employed for technological solutions; however, they have recently been used in a much broader context such as governance design and community problem-solving.  

#HACKTPS 

TPS decided to run a police-led hackathon to look at solutions to crowd-sourcing community problems. The specific problem statement issued for the event was: to develop a solution to increase collaboration between the TPS and Toronto communities. Ryerson University provided the space, food and caffeine for all competitors, as well as an intensive 10-week session for start-up companies as a prize for the winning team. Private sector partners from the steering committee also provided a $10,000 prize for the winning team. The event was promoted on multiple social media platforms via the hashtag #HACKTPS and it attracted over 150 competitors. This was an open competition where anyone could compete. This is an important component of any crowdsourcing competition because it maximizes diversity of thought and therefore, the potential for completely new and innovative solutions.  

When participants arrived, they were required to register and then create teams based on skillsets. Once registered, participants were given a wristband which was colour-coded according to their skillset. They then had to find other participants with different skills until they had four distinct kinds of expertise on their team. People had to make decisions about who they wanted to work with quickly and most people had not met before. This method has had positive impacts in the tech sector in Toronto, as many startups are formed at hackathons by the relationships developed through the creation of teams in this way. Through intense time and task pressures, teams quickly solidified or struggled.  

One value-add for participants of hackathons is that they get to learn new skills and understand challenges facing different sectors. During the event, TPS members acted as competition mentors and all teams were given equal access to them. Such mentorship is key as participants may have a great technology idea but not know the legal or ethical implications of what they are designing. It is important to note that the information provided to participants was never confidential and was consistent with what would be provided in any public setting. This is an important consideration in determining the original hackathon problem statement. This is because the participant/mentor conversations are critical to the design process, including the development of final product specifications. Participants and mentors must be able to share freely, so problems related to covert work would not be suitable for a crowdsourcing competition. 

In order to learn about challenges in technology and community safety, throughout the hackathon there were presentations from various stakeholders. The Chief of Police spoke about innovation and public safety, while the university and private sector stakeholders brought in experts to talk about a wide range of issues related to leveraging technology to create community safety.  

At the end of an intensive 44 hours of non-stop hacking and two rounds of judging by a variety of experts, there was one winner and one runner up team. The winning team was a group of engineering students from another university who created an app that gamified community-enhancing activities with rewards for both individuals and local business associations. The runner-up team was a group of high school students who created a chatbot to link community members with policing knowledge without having to tie up police resources.   

Both of the teams have continued to develop their products. One of the teams has incorporated and is moving substantially closer to a marketable solution. They are currently looking for neighbourhoods in which to pilot their application. Interestingly, the product is more focused on community engagement and volunteerism than it is on a law enforcement solution, so it is expected it will find broad success across non-policing civic divisions. This is beneficial because, given the Canadian procurement rules outlined previously, TPS cannot procure the product given that they were integral in its design.  

The hackathon led TPS to consider technologies to enhance community safety that had not been considered before. For example, various applications of robotic process automation were put forward that are now being leveraged in other projects. Further, as the first police-led hackathon in Canada, it introduced many new people from the tech sector to learn about policing challenges and directly participate in solving community problems, while police were able to access the tech knowledge and experience of over 150 participants.  

Conclusion 

Embedding a front-line police officer in a university start-up incubator program began as an investment in questions. Unlike other projects, this pilot was committed to looking at new methodologies, to reconsider how TPS was dealing with opportunities and challenges. This program showcases how unusual partnerships can lead to significant returns on investment.  As a result of the lessons learned, TPS now includes agile methodology, ideation and design thinking services, prototyping, continued engagement with start-ups, more innovative procurement options, and outcome-based competitions such as hackathons. The tangible and sustainable outcomes that resulted from this initiative include a more effective center of excellence for the development of apps and dashboards for both TPS members and the public, a UX designer for the service, the use of UX methodology both internally and by neighborhood officers for improved community projects, co-creation of technology projects through an invitation to partner procurement model, and the use of hackathons to identify emerging technologies to enhance public safety. Creating a culture of innovation in policing will always remain a challenge and it is important to infuse police organizations with unique external perspectives continuously. In this case, start-ups have many lessons to further innovation in policing.  

Shawna Coxon is a Deputy Chief of Police with the Toronto Police Service, where she has worked in many varied areas of policing for more than 23 years. She has her PhD in Criminal Law from Leicester University in the U.K. and has both published and spoken internationally. She has a passion for policing, innovation, technology and futurism and was the Executive Sponsor of this project. She can be reached by email at shawna.coxon@torontopolice.on.ca and via LinkedIn/Instagram/Twitter @ShawnaCoxon 

Ian Williams is the Manager of Analytics and Innovation with the Toronto Police Service for 12 years. His areas of expertise include business intelligence, data science, crime analytics, strategic analysis, continuous improvement approaches, and innovation. He has two Master’s degrees: one in Geospatial Analysis from Ryerson University, Toronto and another in Public Policy, Administration and Law from York University, Toronto. He was the Project Lead and brainchild of this project. He can be reached by email at ian.williams@torontopolice.on.ca