Is LTA falling for a sunk cost fallacy with its ERP 2.0 OBU that fails to meet promises made when the tender was awarded?

In the wake of Singapore’s rollout of the much-anticipated Electronic Road Pricing (ERP) 2.0 system by the Land Transport Authority (LTA), a closer examination reveals a landscape fraught with discrepancies and unmet promises.

As motorists grapple with the introduction of the bulky, outdated-looking On-Board Unit (OBU), which was originally anticipated to be rolled out starting in 2020, questions loom large over the S$556 million allocated for the project and the fidelity to the original tender specifications.

Leveraging Global Navigation Satellite System (GNSS) technology, the envisioned upgraded system promised equitable charging based on distance travelled, alongside real-time traffic updates and a suite of value-added services accessible through an interactive OBU.

Criticism has been directed at the new on-board units (OBUs) due to their bulky design and inconvenient placement of the cash card reader, such as beneath the glove compartment or in the front passenger footwell.

Many motorists have voiced concerns about whether the LTA genuinely considers user feedback, questioning the rationale behind proceeding with and defending the current design.

Many motorists have also questioned why LTA persists in rolling out the bulky 3-piece OBU design, which they find inconvenient, and raised concerns about why LTA couldn’t introduce a more compact 1-piece design.

Senior Minister of State for Transport Amy Khor, in October 2021, responding to public feedback, had said that the authorities will proceed with the device while continuing to explore improvements.

The decision to proceed with the design, which had already been unpopular then, is said to be due to LTA being contractually bound to the OBU design — said by Dr Khor in Parliament — after awarding the tender in 2016; thus, any alterations would incur additional costs.

However, we have, in a previous article, pointed out that it was Chew Men Leong, the CEO of LTA at the time, who said that drivers could look forward to a “sophisticated, smartphone-sized onboard unit” in his 2016 announcement by LTA.

On top of the promise of a smartphone-sized unit, Chew also said other functions of the new system would be available from day one.

The current device is, however, missing the additional features as touted originally back in 2016 and has not implemented distance-based charging, merely serving as an ERP device and issuing alerts.

These include real-time traffic information for every road user, coupon-less streetside parking, and automatic payment for off-peak car owners who drive during peak periods.

Given that Parking.sg, supposedly the brainchild of Lee Hongyi and his team at Govtech, is now in use, the coupon-less streetside parking is pretty much redundant for the device.

So what about real-time traffic information? Instead of Chew’s touted day-one feature rollout, LTA has stated that the OBU will instead offer features to alert motorists when approaching an ERP-charging location and display real-time traffic incident alerts in the future, even though the device is already three years late.

In an earlier explanation, LTA cited temperature concerns. They noted that in the enclosed space of a car interior, temperatures can soar, often exceeding 50 degrees Celsius, especially when the car is parked under direct sunlight.

LTA further explained that the processing unit of the OBU, akin to a mini-computer, cannot function reliably under such high temperatures. Hence, it cannot be safely mounted on the dashboard area of a car. The footwell area offers lower temperatures, providing a more suitable environment.

The thread on the Hardwarezone forum highlighted a significant concern regarding the device’s inability to meet the basic Grade 4 requirements outlined in the JESD22-A108 standard.

This standard specifies the High-Temperature Operating Life (HTOL) testing criteria for electronic devices. Specifically, Grade 4 of the standard mandates that devices should be able to withstand exposure to temperatures of 70°C for 48 hours or 90°C for 24 hours.

The fact that the device fails to meet these fundamental requirements raises serious questions about its thermal performance and overall reliability, as outlined by the industry-standard testing protocol, especially since the original ERP OBU could be designed as a single unit and placed on the dashboard, in the same operating condition when the upgraded device has to have its processing unit mounted away from the sun.

Now, going back to Chew’s original assessment of the anticipated design of the device back in 2016, could it be that the device that NCS and MHI presented in the 18-month trial was actually smartphone-sized but later revised when applied to the island scale as its previous model was unable to achieve the same results?

If it wasn’t, then why did Chew say that the device was a smartphone-sized device instead of just coming out bluntly to say it would be a three-piece device?

Furthermore, in response to media queries about the photos circulated on mycarforum.com, LTA acknowledged that the single-unit designs were used in a study back in 2016.

Looking at the concept design, I would reckon these single-unit devices would not have incurred the wrath of car users if they had been adopted. But then again, the bulky device might have design flaws, such as its weight and mass, which would have posed a risk to car users in the event of an accident, as the suction cup might not be able to hold the weight of the device, especially with its momentum.

A video showcasing the interaction with the device reveals its unresponsiveness, with many commenting that it resembles mobile tablets from ten years ago. They also highlight the difficulty of inserting the card and locating the device.

But then, if the device did not meet the specifications of the tender, shouldn’t the product be thrown back to the contractor rather than compromising with the contractor as to what they can work with?

It is also puzzling for Dr Khor to say that LTA is committed to the design of an OBU, which consists of 3 units, when nothing about the design was known—or that the unit was sold to LTA as a single-piece, smartphone-sized device—until the design was revealed in 2020.

Attempts to dig for the specifications of the tender have been unsuccessful, as the tender for ERP 2.0 seems to be a closed tender instead of being conducted via GeBiz.

True enough, the tender was offered to the shortlisted companies which took part in the 18-month trial at Woodlands Avenue 12 that concluded in December 2012.

Four teams – Kapsch TrafficCom, MHI Engine System Asia & NCS, ST Electronics (Info-Comm Systems) & IBM Singapore, and Watchdata Technologies & Beijing Watchdata System – were invited by LTA to take part in the trial, each receiving S$1 million in funding.

The trial involved the installation of roadside equipment to facilitate the testing process, while on-road testing will be carried out by the companies to evaluate their solutions’ performance.

According to news reports in 2014, LTA later called a tender in which three companies were shortlisted to develop the second-generation ERP system. The three consortia shortlisted to participate in the tender are NCS & MHI Engine System Asia, ST Electronics (Info-Comm Systems), and Watchdata Technologies & Beijing Watchdata System.

The winning bid of S$556 million from the NCS—a subsidiary under Singtel—and MHI partnership is less than half of the S$1.2 billion submitted by ST Electronics, the other qualified bidder. It is interesting to note that both NCS and ST Electronics are wholly owned by companies that are majority-owned by Temasek Holdings.

As many pointed out, even if the NCS and MHI winning bid was the lowest, it was still $556 million for a flawed device, which no one likes, and at this point, it seems to many that the LTA is just going along with the sunk cost fallacy.

This can be illustrated by LTA’s stance on using smartphones for distance-based ERP charging, as opposed to the bulkier and outdated ERP 2.0 device, which virtually everyone carries.

In rebuttal to the suggestion, Dr Khor stated in November last year that this change might inconvenience motorists, who would need to initiate their mobile apps, enter passcodes, or use biometric authentication each time they travel. This could potentially lead to inadvertent neglect of the ERP process—a complaint that has not yet been heard from motorists, in contrast to the numerous negative responses about the current design.

According to Chew, in February 2016, the NCS-MHI’s S$556 million bid was superior to ST Electronics, and it came within LTA’s budget.

A Redditor commentator wrote, “Back then when ncs won, we were laughing cause they surely can’t come out with something with that bid amount. One of the tender specs dictate it to be able to withstand hot sun and heat just like the original IU. And specs calls for an intelligent obu. To withstand the high heat alone you need a rather exp casing and pcb and the budget they cater completely will not be able to sustain. And how intelligent can it be if the design they came out has only a single line of LED for display purpose. But lucky they won cause now they had to face the brunt of it. Sometimes cheap doesn’t means good.”

Given what we have seen of the finished product, perhaps the old saying stays true: “If it is too good to be true, it probably isn’t.”