Electric Vehicle Data Logger - Cloud Battery/BMS Telematics

• Overall, electric car battery (EVB) sales are to grow from ~1 mn units in 2017 to ~60+ mn in 2040 (CAGR of ~20%)
• Further, electric two-wheelers (bi-cycles, scooters, motorcycles) may grow from 40 mn sold per year in 2017 to 85+ mn in 2040
• China is the #1 market with 70%+ global EVB market share by 2020
• EVs will account for 54% of new car sales by 2040 and 0.5 billion (33%) on-road light duty vehicles
• In 2040, EVs are to drive 67% of new car sales in Europe, 58% in US and 51% in China - today, they already drive 37% in Norway
• Growth is driven by a 70%+ price reduction per kWh for li-ion batteries (2010-17) - with another 70%+ expected by 2030
• 2018 top 5 lithium-ion manufacturers are Panasonic (33%), BYD (18%), LG Chem (17%), Samsung (9%) and Wanxiang (5%)

The BMS is an electronic system that integrates with rechargeable batteries to monitor critical data parameters.

These include e.g. state, voltage, current and temperature.

Based on the data, the BMS performs vital tasks:

• Keeping the battery inside it's safe operating area
• Monitoring & reporting the battery state (SoC, SoH, ...)
• Balancing cells to ensure a similar state of charge
• Prolonging the life of the battery
• Communicating with e.g. chargers or external devices

Example: The role of BMS in charging

In lithium-ion batteries, overcharging can lead to overheating - potentially resulting in catastrophic events. Conversely, discharging the battery below e.g. 5% capacity can lead to permanent capacity reduction. In both cases, the BMS manages the charge to avoid thresholds being passed. An example of a more advanced use of BMS is in "intelligent batteries". Here, the BMS provides data to an "intelligent charger" on the battery's specs, condition and usage history - allowing the charger to perform optimal charging.

In automotive context, the BMS needs to be able to communicate with other sensors and ECUs in the vehicle. As CAN bus is the standard in automotives, it's also the de facto standard for EV batteries. As such, it's possible to record data from the BMS of most EV batteries using a CAN bus data logger. In turn, this also enables the use of IIoT CAN bus data loggers with WiFi for e.g. forklift telematics, AGV telematics and other BMS-data-to-cloud solutions.

Specifically, batteries often rely on the CAN bus protocols SAE J1939 or CANopen - providing data on a range of parameters, e.g. temperature, pack voltage, cell voltage, current, errors, SoC. In general, EVs will require more data and as such, CAN FD will also play a core role in enabling the future growth.

Rise of EVs

As EVs rapidly grow towards 2040, there will be a need for integrating battery telematics into commercial vehicle fleets - across trucks, vans, buses, forklifts, AGVs and more. As this market matures, so will electric vehicle telematics

150+ mn
EVs by 2040
(vs 40 mn in 2017)

Rise of Cloud

Like vehicles, batteries provide tons of data that need storage and fast processing - but with the rise of cloud servers, this becomes increasingly simple & low cost. Cloud computing also enables more advanced BMS methods

$300 bn
public cloud market in 2021
(vs ~150 bn in 2017)

Rise of IoT

Collecting battery data via WiFi / 4G has previously been costly and difficult. However, the rise of the Internet of Things (IoT) now also brings low cost CAN bus IoT devices that easily enable the transfer of the battery big data

5+ bn
B2B IoT devices by 2020
(vs ~2.5 bn in 2017)

Rise of AI

The use of Artificial Intelligence (AI) for EVBs has been impossible on the existing BMS systems - but with the full set of big data in the cloud, next level battery optimization will be possible - and key to stay competitive

50%+ CAGR
in AI enterprise market
($1 bn in '17 to $30+ bn in '25)

Or, as stated in the below article published in the Power and Energy Magazine IEEE (2017):

"The continuing proliferation of and advances in information and communication technologies, development of powerful cloud computing capabilities, and a growing Internet of Things will significantly enhance or even transform the concept of battery management, as modeling and control of thousands of cells in large-scale battery storage will become easier."

— Hu, C. Zou, C. Zhang, Y. Li, "Technological Developments in Batteries: A Survey of Principal Roles, Types & Management Needs"

As mentioned above, most electric vehicles and battery management systems are CAN based. If you are the OEM, you'll also have the decoding rules for scaling the raw CAN bus data to human-readable form.

In some cases, however, you may want to record electric vehicle data - but you you do not have the decoding rules. Below we outline the typical options you have in this scenario:

• Contact the OEM: In some cases you may be able to request a DBC file and/or PDF with decoding rules from the OEM. This is often possible when working with smaller OEMs and specialized EV products. However, you'll rarely be able to get this information for e.g. private electric cars.
• Research online: You may be able to find databases where others have found decoding rules for e.g. electric cars (see below)
• Reverse engineer the data: As a final resort, you can attempt to use a CAN bus sniffer tool to reverse engineer the CAN bus data - though it's difficult if you need a large number of parameters

By researching online you'll often be able to find e.g. DBC files (CAN databases) containing decoding rules for electric car brands. DBC files can be loaded by most CAN software tools for data decoding, including e.g. the free CAN software asammdf for use with the CANedge.

Typically, these contain EV parameters like State of Charge (SoC), State of Health (SoH), battery current, voltage, temperatures, cell voltages, charging status etc. Below we've listed some of the popular open databases:

• OpenDBC: CAN DBC files for brands like Nissan Leaf (2018), Tesla and more
• Tesla Model 3 & Y: DBC file for Tesla Model 3 and Tesla Model Y
• OBD Dash: Proprietary OBD2 PID info on Mitsubishi, Renault, Subaru, Opel, Hummer and more
• Kia Soul: Google sheet containing decoding information for the Kia Soul EV
• Open Garages: Collection of links for decoding info, incl. Prius and more
• EV PID collection: Extended OBD2 PIDs for EVs, incl. Hyundai Kona, Ioniq EV, Kia Niro, Optima, Ray, Soul EV and more
• Nissan Leaf DBC: Nissan Leaf EV DBC files for use in some Leaf EV models (but not 2019+)
• Renault Zoe: CSV with decoding info for the Renault Zoe EV

Note that some of these do not directly provide DBC files with the decoding rules (for example the 'EV PID collection'). This is typically the case if the decoding rules originate from a tool that does not use DBC files for decoding. In such a case, you can however easily create your own DBC file with the relevant signals. For regular CAN data, this is typically straight-forward. For UDS decoding rules, we suggest you check out our github API examples for a sample DBC for the Hyundai Kona EV.

It is important to note that the way you collect electric vehicle data may differ significantly by car brand/model/year.

Broadcast vs requests: In some electric cars, you'll be able to directly connect via the OBD2 connector to access broadcast raw CAN bus data, containing signals like SoC. However, in other cases you'll need to 'request' the data from the electric vehicle via UDS requests (Unified Diagnostic Services) - see e.g. our github API examples. In the API example, we have configured a CANedge2 to send transport protocol UDS requests to a Hyundai Kona EV via the OBD2 connector. We used the EV PID collection above as basis for creating a DBC file with extended multiplexing, for use in decoding the UDS responses via the CANedge Python API.

Rise of UDS: The exposure of UDS requesting rather than raw CAN bus data appears increasingly popular in electric vehicles. For example, the Nissan Leaf models up to 2018 broadcast SoC as part of the raw CAN data available via the OBD2 connector. However, since 2019 the models do not broadcast any CAN bus data via the OBD2 connector - though the Nissan Leaf responds to UDS requests in similar style to the Hyundai Kona.

Contactless reading and EV-CAN: It is also possible to access the raw CAN bus data from an electric vehicle via other methods than the OBD2 connector. For example, you can use a contactless CAN reader to log data directly from the raw CAN low/high wiring harness. Sometimes it is also possible that two CAN buses are used in an electric vehicle, in which case you may log both via a modified OBD2 adapter cable as in our Nissan e-NV200 case study. The 'extra' CAN bus exposed via the OBD2 connector is sometimes referred to as EV-CAN.

Logging data from electric cars can be complex, but you're always welcome to contact us for sparring on your use case.

Once you've reviewed the above, we propose below next steps to get started:

1. Select a 'pilot kit' to get started - e.g. the CANedge2 (8GB) + a suitable adapter cable
2. Design a small proof-of-concept (incl. a very narrow use case)
3. Do a PoC on data collection, conversion & data processing
4. Develop a bare-bones application script in the cloud to solve a basic pain point
5. Scale the PoC to a full segment (BU, country, warehouse, ...) and review
6. If successful, scale up step-by-step by adding more devices and more automated scripts