The Shoplifting Spree at Punggol Central

In the bustling neighborhood of Punggol, a routine shopping trip at the local Sheng Siong supermarket turned into a major security success story. Catherine Tan Li Eng, a 51-year-old saleswoman, allegedly targeted the outlet at Block 622D Punggol Central multiple times in early September 2025. Over seven separate occasions between September 2 and September 11, she made off with a total of 19 bottles of Jacob's Creek red wine, a popular Australian import valued at approximately S$556 in total. Each time, Tan would place the bottles into her reusable shopping bag and proceed to the self-checkout kiosk, deliberately omitting to scan them before leaving the store.

Punggol Central, part of Singapore's growing northeastern residential hub, hosts this Sheng Siong branch as a convenient spot for daily essentials. The supermarket, known for its competitive prices on groceries and household items, attracts a steady stream of families and working professionals. Tan's actions went unnoticed initially, blending seamlessly into the daily crowd of shoppers navigating aisles stocked with fresh produce, meats, and imported wines.

Inventory Discrepancy Sparks Investigation

The scheme unraveled on September 10, 2025, during a standard inventory check at the store. Staff noticed a shortage in the wine section, prompting a review of closed-circuit television (CCTV) footage from the preceding days. The recordings captured Tan's methodical approach: selecting bottles, concealing them in her bag, and bypassing the scanner at self-checkout. Further backtracking revealed not just one or two instances, but a pattern spanning nearly two weeks.

This discovery highlighted a common vulnerability in modern retail—self-checkout systems, which speed up transactions but can be exploited if not monitored closely. In Singapore, where convenience drives consumer behavior, self-checkouts have proliferated in supermarkets like Sheng Siong, NTUC FairPrice, and Cold Storage. However, discrepancies like these underscore the need for layered security measures beyond human oversight.

🛡️ Facial Recognition Alert: Technology Steps In

Once Tan's face was identified from the footage, store staff uploaded her photograph into Sheng Siong's centralized facial recognition system. This artificial intelligence (AI)-powered platform, which matches unique facial features such as the distance between eyes, nose shape, and jawline, cross-referenced her image against prior recordings. The system flagged multiple matching thefts, confirming the extent of her activities.

Sheng Siong's implementation of this technology began in April 2024, initially in about 50 outlets, before expanding to all 83 branches by August 2025. Provided by local tech firm Xinxyong, the system integrates with existing CCTV cameras and a mobile app for staff. When a flagged individual enters any store, an instant alert pings the manager's phone, displaying the person's photo and associated incidents. This proactive approach transforms passive surveillance into real-time intervention.

The Dramatic Interception

True to the system's design, the alert proved decisive on September 11, 2025, around noon. Tan returned to the same Punggol outlet, unaware that her image was now in the database. The retail manager received the notification immediately upon her entry, approached her at the self-checkout, and politely requested to inspect her bags. Inside, three additional unscanned bottles of Jacob's Creek wine were discovered.

Tan initially claimed she had paid for all items, but the omission was clear. Staff recovered the bottles on the spot and summoned the police. Officers from the Punggol Neighbourhood Police Centre arrived promptly, arresting her without further incident. This seamless coordination exemplifies how technology empowers frontline workers, minimizing confrontation while maximizing recovery.

Courtroom Verdict: Eight Days in Jail

Tan's case proceeded to the State Courts, where she pleaded guilty to one amalgamated charge of theft in dwelling under Section 380 of the Penal Code 1871, read with Section 124(4) of the Criminal Procedure Code 2010. Amalgamation combines multiple similar offenses into a single count, escalating the maximum penalty from three years' jail and/or fine to six years' jail and/or fine.

On May 12, 2026, District Judge Brenda Chua sentenced her to eight days' imprisonment, deeming the S$556 value "not insignificant" and the repeated nature indicative of higher criminality. Despite defense pleas for a shorter term citing Tan's role as sole caregiver to her elderly mother and intellectually disabled cousin, and her full restitution of about S$470 for unrecovered bottles, the judge prioritized deterrence. Tan expressed remorse in court, marking a sobering end to her spree.

Sheng Siong's Journey with AI Surveillance

Sheng Siong Group Limited, Singapore's third-largest supermarket chain with over 60 outlets islandwide, adopted facial recognition amid rising retail theft pressures. Founded in 1985 by the Ng brothers, it caters to budget-conscious households with a focus on fresh foods and everyday bargains. Theft per store climbed from 3.09 cases in 2021 to 4.91 in 2023, prompting the tech investment.

The rollout aligns with the Singapore Police Force's (SPF) Shop Theft Awareness for Retailers (STAR) Programme, launched to equip over 1,000 outlets with tools like CCTV and awareness campaigns. Sheng Siong collaborates closely, sharing data to refine algorithms. Beyond theft, the system identifies customers with bounced payments, slashing monthly losses from S$10,000 to S$1,200–S$1,500.

Effectiveness Proven: Detection Rates Soar

Post-implementation metrics are telling. Shoplifters caught per store doubled from four to eight annually between 2023 and 2024, attributed to enhanced visibility rather than crime spikes. While overall reported thefts rose—2,097 cases in the first half of 2025, up 4.2% year-on-year—the per-store uptick at Sheng Siong reflects better reporting and interception.SPF's mid-year brief notes shop theft as the top physical crime, often by youths seeking 'street cred'.

For context, full-year 2025 saw 4,109 cases, a 3% dip from 2024's 4,237, amid economic recovery. Sheng Siong's success has inspired peers, proving AI's return on investment through recovered goods and deterrence.

Singapore's Retail Theft Landscape

Shoplifting remains a persistent challenge in Singapore's compact retail ecosystem. High-density malls and supermarkets like those in heartlands face opportunistic crimes, fueled by factors like inflation, youth boredom, and organized rings. Youth arrests (ages 10-19) surged 41.1% to 271 in H1 2025.

Supermarkets bear the brunt: Cold Storage and Sheng Siong report higher averages per outlet. Economic pressures post-pandemic, including rising living costs, correlate with upticks, though 2025's slight decline signals effective interventions. For detailed trends, the SPF November 2025 release lists multiple Sheng Siong cases among 10 charged.

Privacy Considerations in the Digital Age

While effective, facial recognition raises debates on privacy under Singapore's Personal Data Protection Act (PDPA). Sheng Siong mandates signage at entrances notifying customers of CCTV and AI use, ensuring transparency. Data is stored securely, accessed only for security, and deleted post-resolution.

Critics worry about false positives, biases in AI training data, or mission creep to marketing. Globally, cases like UK misidentifications highlight risks, but Singapore's regulated approach—via Infocomm Media Development Authority (IMDA) guidelines—mitigates them. Retailers balance deterrence with rights, posting notices and limiting retention.CNA's coverage notes no major complaints at Sheng Siong.

Prevention Tips for Retailers and Shoppers

Retailers can adopt hybrid strategies:

• Combine AI with staff training for nuanced checks.
• Optimize store layouts to reduce blind spots.
• Partner with STAR for free resources.
• Monitor high-value items like wines with tags.

Shoppers, deter theft culture by:

• Reporting suspicious activity.
• Using manned checkouts for valuables.
• Understanding penalties—up to 7 years for repeat thefts.

Communities foster vigilance, as seen in heartland watches.

Photo by grainfalls on Unsplash

Looking Ahead: AI's Role in Safer Shopping

Tan's case exemplifies technology's triumph over opportunism, potentially curbing Singapore's 4,000+ annual shop thefts. As AI evolves—integrating behavioral analytics and predictive modeling—retail could see near-zero tolerance. With 2026 projections stable, collaborations like Sheng Siong-SPF pave the way for smarter, safer stores. Shoppers benefit from lower prices, unburdened by shrinkage costs estimated at 1-2% of sales.

Ultimately, this blend of innovation and enforcement reinforces Singapore's reputation for low crime, ensuring Sheng Siong remains a trusted neighborhood staple.