Tire Technology International - March 2024 Issue
OE interview: Goodyear
Graham Heeps 2024-04-06 08:39:19
TARGET IN SIGHT
Goodyear’s Kanwar Bharat Singh talks exclusively to TTI about delivering at last on the intelligent tire’s decades of promise
We have put about 15 million miles on tire sensors and collected a lot of data
Regular readers know that TTI has charted developments in the field of tire intelligence for much of its 31-year existence. Translating the promise of what could be achieved beyond pressure monitoring into a production reality has proved to be a very long game.
As well as driver alerts and wear indicators, we’ve long been told that real-time tire information, directly or indirectly derived, will be crucial to the safety and efficiency of future vehicles.
OE fitments are still thin on the ground, yet progress has been made, notably in fleet monitoring applications, both by tire-mounted sensor (TMS) technologies and maintenance-free, sensorless setups (see also SRI’s Sensing Core, p38) – positive signs for the researchers, managers and marketeers who continue to believe in the value of delivering information about the tire and road surface to the vehicle and beyond.
Tesla TMS
At Goodyear, Kanwar Bharat Singh has been at the forefront of developments in this field for more than a decade. When TTI caught up with Goodyear’s program manager for algorithms and tire intelligence earlier in 2024 there was a lot to discuss. The tire maker’s SightLine product, a tire-agnostic, software-based intelligence solution, has racked up millions of miles in fleet testing and forms the basis of the OE-fit intelligent tires (Goodyear Wrangler Territory RT 285/65 R20s) on the Tesla Cybertruck.
“It’s our first production application,” says Singh. “We strongly believe this will open doors now for other OEMs. There was a lot of reluctance in adopting this new technology – questions about how to manage standards, inventory and logistics. Now we see a lot more interest.”
The Tesla setup features a TMS that feeds tire pressure and temperature information back to the vehicle via Bluetooth. But Singh says the TMS isn’t the essential part of a SightLine implementation.
"For Goodyear, this world of software is new. We are also learning how to build software to put in a production car”
“A key enabler for all of this is RFID,” he explains. “In answer to the question, ‘Do I need a tire sensor?’ I would say ‘yes’ as an algorithm person. If it’s there, a tire sensor does help tremendously. But tire ID is a must-have. The reason we’ve invested in software-only solutions is that it’s very powerful to combine RFID with the software library. The software only needs to know ‘What tires am I working on? Summer, winter, Goodyear, Michelin, Pirelli?’ We have a lot of RFQs now for RFID SOPs. At the plant level, OEMs want to manage the inventory better [with RFID] but now they’re thinking about other use cases. In an ongoing technical project with one OEM, we are giving them the software libraries with which they combine the RFID. I believe this will be the next iteration we will have in the market.”
Singh notes that the establishment of a standard ‘pouch’ for sensor fitment removed one obstacle to further mass-production applications. In the case of the Tesla Cybertruck’s tires, a final production step robotically cleans the tire, adds the sensor and flashes it with the necessary information. The sensor itself is Goodyear’s latest TPMS with a MEMS accelerometer.
“But a lot of the features we have on the car are purely software features,” he stresses. “Tire sensors today are a non-standard piece of hardware. When we try to develop and sell this to an OEM, often the discussion is, would this work on a competitor tire? As we’ve made our product roadmap [for SightLine], we’ve made a lot of these solutions tire agnostic. A hydroplaning model, for example, is a pure software solution that could work on our tire or a competitor’s tire as well.”
ZF cooperation
Hydroplaning mitigation (see Walking on water, above) is one of several ways that Goodyear has developed to use SightLine for active safety; others include friction estimation to optimize automated emergency braking and cruise control, and tire load estimation for stability control. Singh believes OEMs want to mitigate a problem, not just sense it. To determine how to use tire information in low-level, chassis actuation terms, Goodyear in Luxembourg began working with ZF three years ago on ways to improve active safety, in a partnership Singh describes as “a key enabler for us to get to market”.
He admits, “For Goodyear the tire company, this world of software is new. We are also learning how to build software to put in a production car. Having a partner like ZF is great because they do this every day. We can then focus on the core modules. We are not in the business of brakes and stability control systems, and ZF was looking for a tire partner that could give them the sensing.”
“We are trying to give [Gatik] that level of confidence where we can predict the limits of the tire on a given day”
Developing demonstrator cars and common platforms, as well as defining software architectures and coding methodologies for software exchange between SightLine and ZF’s cubiX chassis control software, have all been part of the partnership so far.
“We pull about 20 different signals off the vehicle CAN, including wheel speeds,” Singh explains. “When we work with ZF, we go into the chassis control ECUs to get all that data. We then run the fusion models, combining the tire model with the vehicle feedback to predict hydroplaning and friction levels.
“The tire sensor is a nice-to-have because the physical feedback increases the fidelity of the models, otherwise it’s an open-loop prediction,” he continues. “I can predict if there’s lift-off based on wheel slip, but the tire sensor tells me by how much the tire is deflecting, so I can calculate. Again, the solution could be with or without a sensor, depending on the fidelity [required], and the integration discussion then becomes a cost discussion. On certain AV platforms where they can afford to put a tire sensor, they would go with that. On a regular production car, maybe we just have the software solution to start with.”
Data, data, data
In a bid to prove SightLine’s real-world capability, Goodyear has spent the past few years gathering millions of miles of real-world data from fleet customers and trials. This includes a predictive maintenance program – pressure and wear monitoring – on the autonomous fleet of DiDi, a Chinese rideshare company.
Singh say, “There we have put about 15 million miles [24,000,000km] on tire sensors and collected a lot of data. It’s helped us work with suppliers to make sure the system is robust and master the attachment process on different tires, different sizes. There’s a lot of complexity there. Today we have a system that we feel comfortable pitching to OEMs because it’s well tested in durability terms. With OE projects, the question is often, ‘How many million miles did you run on this?’ And we have those millions of miles to show now.”
"Today we have a system that we feel comfortable pitching to OEMs because it’s well tested in durability terms”
Other trials include paid pilots where TMS has been deployed to collect data on more than 10,000 trucks, streaming data back to the cloud to run analytics and provide feedback. A further ongoing trial is with Gatik’s autonomous trucks, which deliver groceries for Canadian chain Loblaw in the Toronto area.
Singh adds that conversations with Gatik began around predicting the truck’s stopping distance, which led to the need for grip estimates that could be used for the truck’s high-level controller to calculate safe cornering speeds and following distances (see Shopping for solutions, p69).
Back at the lower level of control, Goodyear has also been working with TNO in the Netherlands since 2021, exploring how conventional brake performance can be improved when information about the tire and road surface is fed into the brake controller. The difference between this project and previous industry efforts, Singh says, is that the experiments have been with a physical production car and production-grade ABS, rather than in simulation. That the team cut as much as 1.75m from a car’s braking distance is further evidence that tire intelligence still has the potential to improve road safety when it’s adopted on a wider scale.
New sensor partnership
Finally, Goodyear is tackling the sensors themselves. It currently sources these through strategic partnerships with TPMS suppliers in Germany, which are developing hardware to Goodyear’s specification.
“These systems do need adjustment [for SightLine],” says Singh. “They are not your standard TPMS because of other features – libraries, accelerometers and battery management systems. And even though you’re doing more computing on the chip, you still have to meet the TPMS legislation, which is 6 to 10 years of battery life.”
For the past year, the tire maker has been working with TDK at the start of a longer-term project to bring a new tire sensing system to market.
“TPMS is a volume business,” he observes. “Standard chips sell in the millions. A tire-mounted sensor is nothing but a repackaged rim-mounted system with an accelerometer. The challenge we’ve often had with suppliers is the willingness for them to invest in new TPMS systems without a business case behind it. With TDK we have a strong sensor partner with whom we can start building our own sensing elements. TDK does the fundamental sensors, like an accelerometer, for example, but we would still need to go back to an integrator who would package it in the TPMS form factor and put it in the tire.
“With TDK we are looking at some new sensing technologies and, especially, addressing pain points. For example, in fleet applications we still have use cases around irregular wear conditions – don’t just tell me that I have two millimeters [left], tell me whether I have an issue on the shoulder. With TDK, we are developing new sensing elements that can specifically address some of the technology gaps we have.”
Another topic under investigation is battery life. The more intelligence that’s added to the sensor, the more the battery will be drained, a situation Singh describes as a “bottleneck”. The goal is to maintain the current 6- to 10-year life despite the higher transmission rates of the added safety functions, which require sending data every 5 to 10 seconds for load and friction calculations instead of maybe once a minute for TPMS.
“We are also working on energy harvesting with TDK,” he reveals. “We’re addressing all of those topics that will make [intelligent tires] a saleable package, not a gimmick at a trade show.
WALKING ON WATER
Developing SightLine for active safety applications
Kanwar Bharat Singh explains how SightLine combines with ZF’s chassis controller to mitigate hydroplaning.
“There are two different modules. One of them is purely prediction. I know the state of the tire: let’s say it’s a half-full tire, 2psi lower inflation pressure, it’s lightly loaded, it’s raining. We combine weather data so I can calculate the likelihood of hydroplaning. Then ZF comes in and says, if there’s a high likelihood of hydroplaning, I can adjust my adaptive cruise control. You’ve set it to 65mph [105km/h], I’m going to bring it down to 55mph [89km/h], for example. Once you enter a pool of water and there’s physical hydroplaning, ZF uses the brake and steering actuators to stabilize the car. We are doing the detection for them and telling them on a scale of zero to 100, what the severity of the hydroplaning is in terms of loss of contact. They can then decide what sort of an actuation strategy they want to use to stabilize the car.”
SHOPPING FOR SOLUTIONS
Inside the relationship between Goodyear and Gatik
Following a Canadian trial, autonomous truck maker Gatik plans to implement Goodyear’s SightLine intelligent tire solution in a larger proportion of its North American fleet during 2024.
“I think the main learning we’ve had is Gatik’s clear need for precise information about the potential of the tire on any given day, at any given point in time,” says Goodyear’s Kanwar Bharat Singh. “They drive in snowy and slushy weather but historically AV companies have had a conservative approach and only driven on sunny days. We are trying to give them that level of confidence where we can predict the limits of the tire on a given day so that they can program their perception system and the control module to keep the truck safe.
“Another application we work on with their controls team concerns lane keeping, feeding information into the controller to decide how much steering to put in to keep the truck on the defined path. Cornering stiffness and braking stiffness are tire properties that dictate the response of the tire. On a cold day [versus] on a sunny day, these properties might change by 30% but to date they had these values hard-coded, and the controller’s always trying to correct it. Some of the refinements we are bringing in are advanced tire models using tire sensors, where we are updating these parameters and feeding them to the controller.”
©UKi Media & Events. View All Articles.