Or, what I think about when surveillance cameras watch me.

This is a short reflection essay after observing the state-of-the-art surveillance cameras that have been installed in Geylang as part of the Lamppost-as-a-Platform (LaaP) programme1, the latest iteration of Singapore’s state-managed public surveillance plan.

Uncle points towards another alley and tells us that the upgraded surveillance cameras are down that way, you know, the cameras that “follow you”. So we head there, and come face to face with a black solid sphere mounted high up on a lamppost. As we notice it, so it notices us. With a soft mechanical whirl and a swivel, the gaze of its lens falls onto us. Unlike the older generation of surveillance cameras whose watchful gaze were kept hidden behind tinted glass, this one meets us eye to eye. The message is clear: “I’m watching you.

State-of-the-art surveillance camera installed in Geylang

Uncle is happy that our attention is now focused onto the newer cameras, almost like he’s showing them off. But the cameras do not belong to him. No, they belong to the state. These state-of-the-art surveillance cameras are part of Singapore’s Lamppost as a Platform program2, under the Smart Nation program. These new cameras present unprecedented tracking capabilities, with crowd detection and facial recognition enabled by advancements in computer vision research. Unlike San Francisco, which has banned facial recognition technologies for public surveillance3, these installations reveals the embrace of an AI-powered surveillance infrastructure by Singapore, with the ultimate intention of island-wide deployment in the very near future4.

Most Singaporeans have adopted a kind of ambient acceptance towards the impasse which is state-managed public surveillance5, even as the nation’s obsession to spawn surveillance cameras in public spaces continues to grow6. While the latest manifestation of obsession has sparked privacy concerns, it is also quickly downplayed, falling back to a ‘safety first’ argument1, which aligns neatly with Singapore’s survivalist narrative7.

Still, I could not help but feel a sense of unease towards being so blatantly watched. Even as you walk down the street, the ‘gaze’ has now adopted a gestalt nature, hopping from one camera to the next. And even if you try not to pay attention to them, even if you try to not catch their gaze, you can still hear them watching you, as you pass each lamppost, the soft whirring of spheres turning to cast you in their vision. Contemplating on this unease, I have identified three sources, which I will list as vignettes of thoughts below.

First, what exactly do the cameras see? With fuzzy terms like crowd analytics, it’s hard to really know what is being captured beyond our physical image. Deploying models that seek to predict criminality or detect emotions is tempting in the pursuit of predictive policing8, but AI researchers have warned how such classification systems share the same treacherous grounds as physiognomy, the unscientific study of physical features to determine social and emotional traits9. The opaqueness of these infrastructures of surveillance only leaves us spinning local myths and folklore on what these cameras can see and cannot see. Such anxieties have no place in public space and undermines the sense of security that these cameras are supposed to afford.

An image dataset for predicting criminality from Shanghai Jiao Tong University

Second, it occurred to me that such a network of cameras also constitutes the increasing institutionalisation of neighbourly care10. Surveillance cameras alienate us from the responsibility of keeping watch for our neighbours, as well as undermining the power of first person witnesses, as we value more the bird’s eye view of surveillance footage. As we lament the loss of our ‘kampong spirit’, we also no longer fear the witness of our neighbour, but instead point to the nearest camera and say, ‘Look, government is watching,’ in the same way that we now readily whip out our smartphones to record any scenario that demands samaritan care.

Finally, while I imagine the landscape of surveillance, the areas that fall under the watchful gaze of the programme, I also wonder about the spaces unseen, the blind spots. Media scholar Shannon Mattern highlights that ‘rendering absence’ and ‘mapping erasure’ presents rich grounds for theoretical work11. What then might these unseen spaces reveal about surveillance? Who knows about these spaces and what they are used for? While the government might have such a mapping, accessibility is an issue. Then, what about Uncle? And all the other residents? Certainly if he has a mapping of where the surveillance cameras are, he has, even if unintentionally, constructed a mapping of blind spots. Such maps constitute a tacit knowledge before it is drawn out in paper, a form of power held by residents and by communities. The Simpsons episode ‘To Surveil With Love’ illustrates a possible future.

Screengrab from the Simpsons episode ‘To Surveil With Love’

Singaporeans should rise up from their passiveness towards our increased state of surveillance and begin to hold these infrastructures of surveillance accountable. This extends beyond surveillance cameras which I have discussed here to other surveillance technologies like contact-tracing devices and camera-equipped doorbells. However, such accountability first demands a level of transparency which we are not currently afforded. However, the tacit knowledge of local communities might perhaps serve as initial grounds of evidence to begin such conversations and reduce our dependence on those in power to command information, enabling action from the ground up.


  1. Abdullah, Z., & Cheng, I. (2018, May 14). “Smart lamp posts” in Singapore won’t shine light into people’s lives. CNA. https://www.channelnewsasia.com/news/singapore/smart-lampposts-singapore-wont-shine-light-peoples-lives-10229804  2

  2. Government Technology Agency. (2019). Lamppost as a Platform. GovTech. https://www.tech.gov.sg/scewc2019/laap 

  3. Conger, K., Fausset, R., & Kovaleski, S. F. (2019, May 16). San Francisco Bans Facial Recognition Technology. The New York Times. https://www.nytimes.com/2019/05/14/us/facial-recognition-ban-san-francisco.html 

  4. Tham, I. (2018, October 12). ST Engineering wins $7.5m “smart lamp posts” tender. The Straits Times. https://www.straitstimes.com/tech/st-engineering-wins-75m-smart-lamp-posts-tender 

  5. Tay, T. F. (2020, May 26). S’poreans ready to give up some privacy for safety: IPS study. The Straits Times. https://www.straitstimes.com/singapore/sporeans-ready-to-give-up-some-privacy-for-safety-ips-study 

  6. Sim, R. (2016, September 26). More surveillance cameras as deterrent. The Straits Times. https://www.straitstimes.com/singapore/more-surveillance-cameras-as-deterrent 

  7. Emily, C. H. C. (2019). Survival by Technopreneurialism: Innovation, Imaginaries and the New Narrative of Nationhood in Singapore. Science, Technology and Society, 24(3), 527-544. 

  8. Fussell, S. (2020, June 24). An Algorithm That “Predicts” Criminality Based on a Face Sparks a Furor. Wired. https://www.wired.com/story/algorithm-predicts-criminality-based-face-sparks-furor/ 

  9. Crawford, K. (2021). The Atlas of AI. Yale University Press. 

  10. Illich, I., & Lang, A. (1973). Tools for conviviality. 

  11. Mattern, S. (2021, March 23). How to Map Nothing. Places Journal. https://placesjournal.org/article/how-to-map-nothing/