Achieving safer streets without sacrificing freedom with a trust infrastructure

Imagine a society where children walk safely through their neighbourhoods, play in parks, and explore the world around them.  

A society where no one has to restrict their activities out of fear of crime, and everyone can freely enjoy public spaces. Should this not be one of the fundamental goals of every community? 

For approximately twenty years, I have been conducting research and field experiments on community-based security camera systems in pursuit of this vision.  

I am now proposing a new form of social infrastructure that uses existing streetlight networks to provide continuous monitoring of public spaces, while strictly governing how the resulting information can be used. 

This proposal has two important features.  

First, the essential technologies already exist, and implementation maynot require the expensive, specialised systems that many people imagine.  

Second, technological feasibility and public acceptance are entirely different challenges to tackle. 

Continuously monitoring public spaces using streetlights and smartphone technology 

Many cities already have security cameras, but their coverage tends to be fragmented. 

Imagine a missing child appearing on a camera at an intersection. If there is no camera covering the next section of the street, investigators may have no way of knowing which direction the child went. 

If neighbouring cameras had overlapping fields of view, allowing a person's movements to be followed continuously, search and rescue capabilities could change dramatically.  

The same capability could help trace escape routes after crimes and potentially deter criminal activity. 

This would represent a qualitative transformation in public safety capabilities, instead of simply increasing the number of cameras. 

I call such an environment a Fully Monitored Public Space (FMPS). 

Streetlights are already widely installed along roads.

Meanwhile, modern smartphones combine sophisticated cameras, computing power, communication capabilities, and data storage in compact, mass-produced devices. 

By integrating several small cameras and smartphone-equivalent processing and communication capabilities into existing streetlights, and designing overlapping camera views, it may be possible to achieve continuous visual coverage of public spaces. 

Another advantage is the ability to share existing infrastructure: streetlight poles, electricity supplies, lighting, installation arrangements, and maintenance systems.  

Introducing these capabilities when streetlights are routinely replaced could further reduce additional infrastructure costs. 

When local government officials think about comprehensive public space monitoring, many still think of expensive specialised equipment and large centralised systems.  

The possibility of transforming existing road infrastructure into a new public safety network using streetlights and smartphone technology is still insufficiently recognised. 

What I am proposing combines inexpensive, mass-produced technologies with existing streetlight networks to create public safety capabilities that are qualitatively different from what we have today. 

This opportunity is potentially available to any local authority with a streetlight network. 

But practical implementation will require testing. Physical obstructions, nighttime image quality, communication reliability, maintenance, power supply, and total costs all need to be evaluated.  

The integrated system itself remains at the development and demonstration stage. 

Nevertheless, its essential technological components already exist. This offers local governments an important new option worth exploring. 

Lessons from twenty years of community safety research 

My work dates back approximately twenty years to the e-JIKEI Network Project, a community safety initiative in Japan that recreates traditional neighbourhood watch systems using modern, low-cost technology. 

The project proposed a system in which citizens and local communities could install and manage their own security cameras to contribute to public safety.  

We conducted field experiments in Japan in cooperation with local governments, police organisations, and community members. 

In one experiment, we installed 11 cameras on streetlight poles. The recorded footage was stored in encrypted form.  

When incidents occurred, the footage was provided to the police following a decision by its owner, the local Parent-Teacher Association (PTA). 

These experiences led me to recognise two fundamental challenges. 

The first is that simply having cameras is not enough. When coverage is fragmented, the information available for searching and tracking people remains limited. 

The second is that greater monitoring capabilities increase public concerns about the misuse of personal information. 

Even if technology significantly improves public safety, a project may collapse if citizens oppose it because they fear surveillance, or local authorities don’t propose the project in the first place because they anticipate opposition. 

The answer is not to abandon either privacy or safety. We need to develop stronger monitoring capabilities and stronger mechanisms for governing their use at the same time. 

Protection against hackers is not enough to earn public trust 

When discussing privacy protection in security camera systems, attention often focuses on encryption, preventing unauthorised access, and cybersecurity. 

These technologies have advanced considerably. Yet another equally important challenge remains. 

What happens if a public official with legitimate access privileges misuses those privileges? 

Such capabilities could potentially be misused by individual officials with legitimate access, for example, to track political opponents or monitor acquaintances. 

This involves an authorised insider using legitimate access privileges for an illegitimate purpose. 

Of course, laws, guidelines, access controls, and activity logs already provide important safeguards. However, if surveillance capabilities expand far beyond what is possible today, will these measures alone be enough to earn public trust? 

I doubt that assurances such as "Our staff will use the information appropriately" or "We will follow the guidelines" will be sufficient. 

What is needed is a mechanism that does not depend solely on the goodwill of officials.  

Misuse should be detectable, responsibility should be identifiable, and the consequences of misconduct should outweigh any expected benefits. 

Equally importantly, citizens themselves must be convinced that these safeguards are trustworthy. 

Building new infrastructure for trust 

One possible response to this challenge is a framework I have been developing called Verifiable Record of artificial intelligence (AI) Output (VRAIO). 

Its basic idea is to govern the release and use of information processed by AI through established rules, independent recording, auditing, and accountability. 

Consider a case in which authorities need access to camera footage to search for a missing child. 

The person requesting the information would declare who is making the request, why the information is needed, what information is required, and to whom it will be disclosed. 

The request would then be checked for formal compliance with rules established in advance through legitimate democratic procedures.  

An independent recording institution would preserve tamper-resistant records of the declared purpose, scope, and decision to permit or deny the information release. 

Importantly, the independent recorder would not need access to the footage itself. 

Whether the declared purpose was genuine would be examined through subsequent independent audits. 

If violations can be detected, responsibility established, and appropriate sanctions imposed, the incentive to misuse information can be reduced. 

This framework cannot prevent every information leak or eliminate all forms of misconduct.  

However, it could provide an important foundation for making information use verifiable and discouraging misuse by authorised insiders. 

I regard mechanisms of this kind as trust infrastructure. 

My research argues that while measurement is essential for understanding and improvement, applying it to society requires something further upstream: trust infrastructure. 

Trust infrastructure supports trust. Trust enables social acceptance. Social acceptance, in turn, enables the legitimate use of society's measurement capabilities. 

Without trustworthy governance, even highly beneficial measurement technologies may remain socially unacceptable and therefore unused. 

This idea extends beyond public space cameras. It may apply to many forms of public service that rely on AI  and the collection or analysis of information. 

What local governments should do next 

The integrated FMPS and VRAIO system I am proposing has not yet been deployed by a municipality. 

My recently published research presents the conceptual framework, governance design, and approach for evaluating social acceptance.  

The next step is to develop and demonstrate an integrated system in a limited area. 

For example, a pilot project could focus on a school route and its surrounding streets.  

Streetlight-mounted cameras could provide continuous visual coverage, while information access and disclosure would be governed through independently recorded and audited procedures. 

Such a demonstration should evaluate more than technical performance and installation costs. 

It should examine whether residents feel safer, whether monitoring causes people to avoid public spaces or restrict their behaviour, and whether governance mechanisms can effectively discourage misuse. 

Citizens should also be involved in designing the rules. The arrangements should allow for modification or termination if the results or public concerns justify doing so. 

This represents more than an opportunity for pioneering local governments to install new equipment. 

It is an opportunity to establish new ways of delivering public services by integrating roads, lighting, communications, AI, information governance, and independent auditing.  

Early participants could also help shape common technical specifications and future governance arrangements. 

Existing streetlight networks could evolve from infrastructure that simply illuminates roads into infrastructure that supports both public safety and individual freedom. 

Society could develop not only greater capabilities to observe and understand public spaces, but also stronger mechanisms to ensure that those capabilities are used legitimately. 

The essential technologies needed to pursue such a society are already in our hands. 

I hope that local governments, businesses, researchers, and citizens will work together to take the first step towards demonstrating this approach. 

If these efforts succeed and spread internationally, they could help establish a new standard for the cities and communities of the future—and potentially become one of the defining achievements in social technology and public infrastructure of this century. 

----------------------------- 

Yusaku Fujii is a Professor at the Graduate School of Science and Technology, Gunma University, Japan. Trained in precision measurement, he contributed to research toward the redefinition of the kilogram.  

For around twenty years he has led the e-JIKEI Network Project, conducting community safety field experiments with local governments, police and residents, and he serves as Representative Director of the NPO e-JIKEI Network Promotion Institute.  

His current research focuses on AI output governance and trust infrastructure for public services.