arxiv: 2604.17991 · v1 · submitted 2026-04-20 · 📡 eess.SY · cs.SY

Recognition: unknown

EcoTIM: Fuel-saving multi-brand tillage with ISO 11783 TIM

Ruben Hefele , Timo Oksanen

Authors on Pith no claims yet

Pith reviewed 2026-05-10 04:18 UTC · model grok-4.3

classification 📡 eess.SY cs.SY

keywords fuel optimizationtillageISOBUSTractor Implement Managementmulti-branddiesel consumptiondistributed controlCVT tractors

0 comments

The pith

EcoTIM lets tractor and implement jointly minimize tillage fuel use by exchanging only two efficiency scalars over ISOBUS.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper presents EcoTIM, a distributed optimization method in which the tractor and tillage implement cooperate in real time to reduce fuel consumption per hectare. The tractor fuses its internal efficiencies into one scalar value plus its derivative with respect to speed and sends both to the implement at the standard 100 ms rate. The implement combines these scalars with its own draft force model to compute the fuel-consumption gradient, then commands the optimal speed and acceleration back to the tractor. Only these two values are exchanged, so neither party reveals proprietary models, enabling multi-brand and plug-and-play operation. The authors propose three new messages plus a backward-compatible extension to the TIM speed command for standardization in ISO 11783 and validate the idea with a simulation on six scenarios and a 1 km track.

Core claim

The tractor ECU computes and broadcasts a combined efficiency scalar together with its derivative with respect to vehicle speed; the implement ECU evaluates the fuel-consumption gradient from these scalars and its analytically known draft force model, then issues optimal speed and acceleration commands back through an extended TIM interface, achieving real-time minimization of fuel per hectare without disclosure of internal subsystem models.

What carries the argument

The two-scalar broadcast of combined efficiency and its speed derivative, which permits the implement to evaluate the fuel-consumption gradient while preserving proprietary models on both sides.

If this is right

The architecture is inherently multi-brand and plug-and-play because only two scalars are exchanged and no proprietary models are disclosed.
Three new messages plus a backward-compatible byte extension to the existing TIM speed command are required and proposed for inclusion in the ISO 11783 standard.
The acceleration command supplies feed-forward information that improves tractor transient response over speed-only commands.
Simulation across six tillage scenarios on a spatially varying 1 km track provides initial evidence that the gradient-based optimization reduces fuel use.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The same two-scalar exchange pattern could be applied to other field operations such as seeding or spraying where speed also affects implement efficiency.
Because the data exchange is minimal and uses the existing CAN bus, manufacturers could add the feature through software updates without new hardware.
Spatially aware versions could incorporate GPS or soil-sensor data to adjust the target speed along the field.
The approach reduces the attack surface for cybersecurity threats relative to methods that would require full model sharing.

Load-bearing premise

The implement ECU can accurately compute the true fuel-consumption gradient by fusing the two received efficiency scalars with its analytically known draft force model.

What would settle it

A controlled field test in which measured fuel consumption per hectare shows no reduction when EcoTIM commands are active compared with standard operation, or in which the implement's computed gradient deviates substantially from observed fuel use because the draft force model or received scalars are inaccurate.

Figures

Figures reproduced from arXiv: 2604.17991 by Ruben Hefele, Timo Oksanen.

**Figure 3.** Figure 3: Transmission efficiency versus vehicle speed at four [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗

**Figure 2.** Figure 2: Engine be map (Willans line, crankshaft-referenced after PTO correction with ηPTO = 0.93). Contour lines show constant be in g/(kW h). Orange dots: DLG measured operating points rescaled to crankshaft power, labelled with the corrected be. Thick line: full-load torque envelope. Dashed lines: constant-power hyperbolas. Transmission efficiency. An analytical model based on the Kress [15] power-split framewo… view at source ↗

**Figure 4.** Figure 4: Tractive efficiency versus wheel slip for four soil cone [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗

**Figure 5.** Figure 5: Draft force versus speed for all six tillage imple [PITH_FULL_IMAGE:figures/full_fig_p008_5.png] view at source ↗

**Figure 6.** Figure 6: Virtual 1 km test track: cone index (left axis) and elevation (right axis) versus position. Eight soil and terrain zones with smooth transitions. 800 kPa to 1300 kPa) are each assigned to two zones, with firmer soils on hilltops and softer soils in the valley, reflecting the natural drainage pattern. An elevation profile with two hills provides grade variations of up to ±8 % (approximately 10 m total eleva… view at source ↗

**Figure 7.** Figure 7: EcoTIM 1 km field pass for all six scenarios. Each subplot shows vehicle speed (left axis, km/h, black) and fuel rate (right axis, L/h, blue) versus track position. Solid lines: EcoTIM enabled. Dotted lines: baseline at the scenario-specific operator speed (actual speed shown, which drops where the engine cannot sustain the setpoint). The baseline speed for each scenario is indicated in the subplot title. … view at source ↗

**Figure 8.** Figure 8: Fuel consumption (L/ha, linear axis) versus time per hectare (min/ha, logarithmic axis) for all six scenarios. Dashed lines with circles: baseline at nine constant speed setpoints (4 km/h to 12 km/h). Stars: EcoTIM optimiser result. Each scenario is colour-coded and labelled. Heavy implements (S1 to S4) occupy the upper right; light, wide implements (S5, S6) occupy the lower left. 8.1 %), where the operato… view at source ↗

**Figure 9.** Figure 9: Energy balance per hectare for all six scenarios, comparing the scenario-specific baseline speed with EcoTIM enabled. [PITH_FULL_IMAGE:figures/full_fig_p011_9.png] view at source ↗

read the original abstract

Tillage operations account for a large share of on-farm diesel consumption, yet the fuel efficiency of the combined tractor-implement system is not optimised in current practice. Modern continuously variable transmission (CVT) tractors minimise engine fuel consumption internally, but they treat the implement as an unknown load and do not account for the effect of vehicle speed on implement draft force. This paper presents EcoTIM, a distributed fuel-optimisation concept in which the tractor and tillage implement cooperate through the extended ISO 11783 (ISOBUS) Tractor Implement Management (TIM) interface to minimise fuel consumption per hectare in real time. In the EcoTIM concept, the tractor Electric Control Unit fuses its internal engine, transmission, and traction efficiencies into a single combined efficiency value and its derivative with respect to vehicle speed, and broadcasts both to the implement at the standard 100 ms CAN bus cycle. The implement ECU combines these two received scalars with its own analytically known draft force model to evaluate the fuel-consumption gradient, and commands the optimal speed, and as a novel TIM extension, the desired acceleration, back to the tractor. Because only two scalar values are exchanged and neither party discloses proprietary subsystem models, the architecture is inherently multi-brand and plug-and-play. The required data exchange is realised with three new messages and one backward-compatible byte-level extension to the standard TIM speed command, and this paper proposes that these messages are standardised within ISO 11783. The acceleration command enables feed-forward torque and CVT ratio planning on the tractor side, improving transient response compared with speed-only TIM commands. This paper also contains a proof-of-concept simulation with six tillage scenarios and a spatially varying 1km test track for initial concept validation.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

EcoTIM gives a workable distributed fuel optimization over ISOBUS with two scalars and an acceleration extension, but the gradient calculation misses the load-sensitivity cross term when draft changes with speed.

read the letter

The paper's core move is to let the tractor broadcast a combined efficiency scalar and its partial with respect to speed, then let the implement use its own draft model to pick the next speed and acceleration command. That keeps both sides from exposing internal maps and fits inside the 100 ms TIM cycle. The acceleration command is the part that actually adds something new for transient CVT planning, and the three proposed messages plus the backward-compatible byte extension are spelled out clearly enough that a standards group could pick them up directly. The six-scenario simulation on a 1 km track with spatially varying soil shows the loop can close without blowing the cycle time, which is useful evidence for a first cut. The weak point is exactly the one the stress-test flags. Once speed changes, draft force changes, load torque changes, and the total fuel derivative picks up a cross term involving how efficiency itself varies with load. The tractor only sends the two scalars evaluated at the current point; nothing in the description indicates it also sends a load-sensitivity term or that the implement is expected to reconstruct it. If the simulation treats efficiency as load-independent, the reported savings are optimistic. The abstract and the concept description do not show the full equations, so it is impossible to check whether that term was included or approximated away. This is a standards-oriented proposal aimed at agricultural machinery engineers who already work with ISO 11783. A reader who needs a concrete architecture for multi-brand fuel trimming will find the message layout and the acceleration addition worth the time. I would send it to peer review. The idea is grounded in an existing standard, the simulation is reproducible in principle, and the gap on the gradient is fixable with either more transmitted data or a clearer justification that the cross term can be neglected. A referee can ask for the exact derivative derivation and for field data that tests the load-dependence assumption.

Referee Report

2 major / 2 minor

Summary. The paper proposes EcoTIM, a distributed real-time fuel optimization architecture for tillage that extends the ISO 11783 TIM interface. The tractor ECU fuses its engine/transmission/traction efficiencies into a single scalar ε and its partial derivative ∂ε/∂v (evaluated at the current operating point) and broadcasts both at the standard 100 ms CAN cycle. The implement ECU combines these with its analytically known draft-force model F_d(v) to compute the fuel-consumption gradient, then commands the optimal speed and (as a novel extension) the desired acceleration back to the tractor. Only two scalars are exchanged, preserving proprietary models and enabling multi-brand plug-and-play operation. The required messages are defined and proposed for standardization; a proof-of-concept simulation with six tillage scenarios on a 1 km spatially varying track is included for initial validation.

Significance. If the gradient computation is accurate, EcoTIM could deliver measurable fuel savings per hectare in tillage without requiring either party to disclose subsystem models, directly addressing a major source of on-farm diesel use. The backward-compatible byte-level extension and 100 ms timing respect existing ISOBUS constraints, while the acceleration command enables feed-forward CVT planning. The simulation provides an initial existence proof, but the absence of quantitative error metrics or baseline comparisons limits the strength of the validation.

major comments (2)

[§3] §3 (EcoTIM concept description) and the gradient-evaluation step: the manuscript states that the implement computes the fuel-consumption gradient from the two received scalars (ε and ∂ε/∂v) plus its draft model F_d(v). However, because F_d changes with commanded speed, the load torque on the tractor also changes; the total derivative therefore contains the cross term (∂ε/∂load)·(d load/dv). No such term is transmitted, and the paper neither states the assumption that ε is load-independent nor provides a justification or reconstruction method for the missing partial. This directly affects the central claim that the two-scalar exchange yields accurate optimization.
[§5] §5 (proof-of-concept simulation): the six tillage scenarios and 1 km track are described, but the reported results contain no quantitative error metrics (e.g., fuel-consumption deviation from optimum, sensitivity to the neglected cross-term), no comparison against a speed-only TIM baseline, and no Monte-Carlo or parameter-sweep analysis. Without these, it is impossible to assess whether the architecture actually achieves the claimed fuel savings under realistic load variations.

minor comments (2)

[§3.1] Notation for the combined efficiency ε and its derivative is introduced without an explicit equation defining how the three subsystem efficiencies are fused; a compact equation would improve clarity.
[§4] The paper proposes three new messages and one byte-level extension but does not include the proposed CAN message IDs or payload layouts; these should be added as an appendix or table for implementers.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive and insightful comments on our manuscript. We address each major comment below, providing clarifications on the modeling assumptions and committing to enhancements in the simulation validation to strengthen the paper.

read point-by-point responses

Referee: [§3] §3 (EcoTIM concept description) and the gradient-evaluation step: the manuscript states that the implement computes the fuel-consumption gradient from the two received scalars (ε and ∂ε/∂v) plus its draft model F_d(v). However, because F_d changes with commanded speed, the load torque on the tractor also changes; the total derivative therefore contains the cross term (∂ε/∂load)·(d load/dv). No such term is transmitted, and the paper neither states the assumption that ε is load-independent nor provides a justification or reconstruction method for the missing partial. This directly affects the central claim that the two-scalar exchange yields accurate optimization.

Authors: We appreciate the referee's identification of this subtlety in the total derivative. In the EcoTIM architecture, the tractor ECU computes ε and ∂ε/∂v at the instantaneous operating point, which already incorporates the current load torque from the implement. The partial ∂ε/∂v is obtained by varying speed while holding load fixed (via internal model or short-term estimation). The cross term is treated as a higher-order effect under the local linearization assumption for small speed adjustments around the current point. We acknowledge that §3 does not explicitly articulate this assumption or discuss its validity range. In the revision we will add a clear statement of the assumption, justify it via the smoothness of typical efficiency maps and the small commanded speed deltas, and note that the two-scalar exchange still yields a first-order accurate gradient for the distributed optimization. revision: yes
Referee: [§5] §5 (proof-of-concept simulation): the six tillage scenarios and 1 km track are described, but the reported results contain no quantitative error metrics (e.g., fuel-consumption deviation from optimum, sensitivity to the neglected cross-term), no comparison against a speed-only TIM baseline, and no Monte-Carlo or parameter-sweep analysis. Without these, it is impossible to assess whether the architecture actually achieves the claimed fuel savings under realistic load variations.

Authors: We agree that the current §5 results are primarily illustrative and would benefit from quantitative support. In the revised manuscript we will augment the simulation section with: (i) explicit fuel-consumption deviation metrics relative to a centralized optimum that has full model knowledge; (ii) a direct comparison against a speed-only TIM baseline to quantify the benefit of the acceleration command; (iii) a sensitivity study that varies the load-dependence of ε to bound the impact of the neglected cross-term; and (iv) a parameter sweep over soil resistance and implement parameters to demonstrate robustness. These additions will be presented with tables and additional figures while preserving the proof-of-concept nature of the study. revision: yes

Circularity Check

0 steps flagged

No circularity: independent efficiency scalars and draft model remain separate inputs

full rationale

The paper's core architecture fuses tractor efficiencies into two scalars (combined efficiency and its speed derivative) and passes them to the implement, which then applies its own analytically known draft-force model to compute the fuel gradient. This separation is explicitly stated as enabling multi-brand operation without model disclosure. No step reduces the claimed gradient computation to a fitted parameter from the same source, a self-citation chain, or a redefinition of inputs. Validation occurs via external simulation scenarios rather than internal re-use of the result. The skeptic concern addresses model completeness (missing load-sensitivity cross-term), which is a correctness issue, not a circularity reduction.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on the implement possessing an accurate analytical draft force model and the tractor being able to fuse its efficiencies into usable scalars without model disclosure; these are domain assumptions without independent evidence supplied.

axioms (1)

domain assumption The implement has an analytically known draft force model
Required to compute the fuel-consumption gradient from the two received scalars.

pith-pipeline@v0.9.0 · 5611 in / 1182 out tokens · 46928 ms · 2026-05-10T04:18:09.943939+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

26 extracted references · 1 canonical work pages

[1]

K. T. Renius,Fundamentals of Tractor Design. Cham, Switzerland: Springer, 2020

2020
[2]

ASAE D497.7 MAR2011 – agricultural machin- ery management data,

American Society of Agricultural and Biological Engi- neers, “ASAE D497.7 MAR2011 – agricultural machin- ery management data,” ASABE, St. Joseph, MI, USA, 2011

2011
[3]

Traction and tractor perfor- mance,

F . M. Zoz and R. D. Grisso, “Traction and tractor perfor- mance,” inASAE Distinguished Lecture Series, No. 27. St. Joseph, MI, USA: ASAE, 2003

2003
[4]

Traktionsmanagement für Traktoren,

B. Pichlmaier, “Traktionsmanagement für Traktoren,” Ph.D. dissertation, Technische Universität München, Munich, Germany, 2012. [Online]. Available: https:// mediatum.ub.tum.de/1098891

work page arXiv 2012
[5]

Modellbasierte Berechnung von Kraftstoffverbräuchen landwirtschaftlicher Verfahrensketten,

L. Trösken, A. Meiners, L. Frerichs, and S. Böttinger, “Modellbasierte Berechnung von Kraftstoffverbräuchen landwirtschaftlicher Verfahrensketten,”Landtechnik – Agricultural Engineering, vol. 75, no. 4, pp. 278–299, 2020

2020
[6]

EKoTech – A holistic approach to reduce CO 2 emis- sions of agricultural machinery in process chains,

M. Decker, L. Frerichs, S. Böttinger, and E. Remmele, “EKoTech – A holistic approach to reduce CO 2 emis- sions of agricultural machinery in process chains,” in SAE Technical Paper 2017-01-1929. SAE Interna- tional, 2017

2017
[7]

ISOBUS – the open high-speed network that enables smart implements: How it was, is, and will be,

T. Oksanen and H. Auernhammer, “ISOBUS – the open high-speed network that enables smart implements: How it was, is, and will be,” inASABE Distinguished Lec- ture Series No. 43. St. Joseph, MI, USA: American Society of Agricultural and Biological Engineers, 2021

2021
[8]

ISO 11783-1:2017 – tractors and machinery for agriculture and forestry – serial control and communications data network – part 1: General standard for mobile data com- munication,

International Organization for Standardization, “ISO 11783-1:2017 – tractors and machinery for agriculture and forestry – serial control and communications data network – part 1: General standard for mobile data com- munication,” ISO, Geneva, Switzerland, 2017

2017
[9]

SAE J1939 – serial control and com- munications heavy duty vehicle network,

SAE International, “SAE J1939 – serial control and com- munications heavy duty vehicle network,” SAE Interna- tional, Warrendale, PA, USA, 2023

2023
[10]

ISO 11783-7:2015 – tractors and machinery for agricul- ture and forestry – serial control and communications data network – part 7: Implement messages application layer,

International Organization for Standardization, “ISO 11783-7:2015 – tractors and machinery for agricul- ture and forestry – serial control and communications data network – part 7: Implement messages application layer,” ISO, Geneva, Switzerland, 2015

2015
[11]

SAE J1939-71 – vehicle application layer,

SAE International, “SAE J1939-71 – vehicle application layer,” SAE International, Warrendale, PA, USA, 2023

2023
[12]

ISO 11783-9:2012 – tractors and machinery for agriculture and forestry – serial control and communications data network – part 9: Tractor ECU,

International Organization for Standardization, “ISO 11783-9:2012 – tractors and machinery for agriculture and forestry – serial control and communications data network – part 9: Tractor ECU,” ISO, Geneva, Switzer- land, 2012

2012
[13]

Fendt 514 Vario S4: DLG PowerMix Datasheet,

DLG e.V., “Fendt 514 Vario S4: DLG PowerMix Datasheet,” DLG Test Center, Groß-Umstadt, Germany, Tech. Rep. 2016-00697, Oct. 2016, test performed on AGCO Fendt GmbH tractor; includes transport and field work power, fuel, and urea consumption measurements. Accessed April 17, 2026. [Online]. Available: https://www.dlg-test.de

2016
[14]

Guzzella and C

L. Guzzella and C. H. Onder,Introduction to Modeling and Control of Internal Combustion Engine Systems, 2nd ed. Springer Berlin Heidelberg, 2010

2010
[15]

Hydrostatic power-splitting transmissions for wheeled vehicles: Classification and theory of oper- ation,

J. H. Kress, “Hydrostatic power-splitting transmissions for wheeled vehicles: Classification and theory of oper- ation,” inSAE Technical Paper 680549. SAE Interna- tional, 1968

1968
[16]

Methode zur Analyse und Bewertung von stu- fenlosen Traktorgetrieben mit mehreren Schnittstellen,

B. Reick, “Methode zur Analyse und Bewertung von stu- fenlosen Traktorgetrieben mit mehreren Schnittstellen,” Ph.D. dissertation, Karlsruher Institut für Technologie (KIT), 2018

2018
[17]

Rechnerische Ver- lustleistungsanalyse von stufenlosen Getriebesystemen am Beispiel des ZF Terramatic11,

J. Ziegler, G. Bailly, and J. Pohlenz, “Rechnerische Ver- lustleistungsanalyse von stufenlosen Getriebesystemen am Beispiel des ZF Terramatic11,” inVDI-Berichte 2226. Düsseldorf, Germany: VDI-Verlag, 2014, pp. 423–431

2014
[18]

Verluste und Wirkungsgrade bei Traktor- getrieben,

H. Reiter, “Verluste und Wirkungsgrade bei Traktor- getrieben,” Ph.D. dissertation, Technische Universität München, Munich, Germany, 1990

1990
[19]

Traction prediction equations for bias ply tires,

W. W. Brixius, “Traction prediction equations for bias ply tires,” inASAE Paper No. 87-1622. St. Joseph, MI, USA: ASAE, 1987

1987
[20]

Ahokas and T

J. Ahokas and T. Oksanen,Maamekaniikka, 2nd ed., ser. Julkaisuja / Maataloustieteiden laitos. Helsingin yliopisto, Maataloustieteiden laitos, 2015, no. 40

2015
[21]

Online con- figurable tractor implement models,

P . Rößler, T. Kautzmann, and M. Geimer, “Online con- figurable tractor implement models,”Landtechnik – Agri- cultural Engineering, vol. 67, no. 4, pp. 247–250, 2012

2012
[22]

Local navigation and ob- stacle avoidance for an agricultural tractor with nonlinear model predictive control,

R. Soitinaho and T. Oksanen, “Local navigation and ob- stacle avoidance for an agricultural tractor with nonlinear model predictive control,”IEEE Transactions on Con- trol Systems Technology, vol. 31, no. 5, pp. 2043–2054, 2023

2043
[23]

TCD 4.1 L4 (Agri) – product data sheet,

DEUTZ AG, “TCD 4.1 L4 (Agri) – product data sheet,” https://www.deutz.com/de/produkte/engine-finder/ engine-detail/tcd-41-l4-agri/, 2024, 4-cylinder 4.0 L Stage IV/V diesel engine, rated 126 kW at 1900 rpm, minimumb e = 208g/kWh. Engine family of the Fendt 514 Vario S4 (Deutz 4038 cm 3). Accessed April 16, 2026

2024
[24]

Darmstadt: Kuratorium für Technik und Bauwesen in der Landwirtschaft, 2018

Kuratorium für Technik und Bauwesen in der Land- wirtschaft (KTBL),Faustzahlen für die Landwirtschaft, 15th ed. Darmstadt: Kuratorium für Technik und Bauwesen in der Landwirtschaft, 2018

2018
[25]

Closed loop control over ISO 11783 net- work – challenges of plug-and-play,

T. Oksanen, “Closed loop control over ISO 11783 net- work – challenges of plug-and-play,” inIFAC Proceed- ings Volumes (IFAC-PapersOnline), vol. 43, no. 26, 2010, pp. 169–174

2010
[26]

Machine vi- sion for soil roughness measurement and control of tillage machines during seedbed preparation,

P . Riegler-Nurscher, G. Moitzi, J. Prankl, J. Huber, J. Karner, H. Wagentristl, and M. Vincze, “Machine vi- sion for soil roughness measurement and control of tillage machines during seedbed preparation,”Soil and Tillage Research, vol. 196, p. 104351, 2020. Technical University of Munich Professorship of Agrimechatronics 14 ISBN:978-3-911430-14-2

2020